Comparative Genomics of NBS Domain Genes: Evolutionary Insights, Methodological Advances, and Applications in Disease Resistance

Abstract

This article provides a comprehensive synthesis of comparative genomic studies on Nucleotide-Binding Site (NBS) domain genes, the largest class of plant disease resistance (R) genes. We explore the remarkable diversification and dynamic evolutionary patterns of NBS-LRR gene families across diverse plant lineages, from asparagus and Rosaceae to Nicotiana and Apiaceae species. The review details established and emerging bioinformatics methodologies for genome-wide identification and classification of NBS genes, addressing common analytical challenges and optimization strategies. We further examine functional validation approaches and comparative frameworks that bridge genomic findings with disease resistance phenotypes, highlighting how these insights are being leveraged to understand susceptibility mechanisms and inform crop improvement programs. This resource is tailored for plant scientists, genomic researchers, and crop development professionals seeking to harness NBS gene diversity for enhancing plant immunity.

The Plant Immune Repertoire: Diversity and Evolution of NBS Gene Families

The nucleotide-binding site-leucine-rich repeat (NBS-LRR or NLR) gene family constitutes a cornerstone of the plant innate immune system, encoding intracellular receptors that confer resistance to diverse pathogens through effector-triggered immunity (ETI) [1] [2]. The architectural diversity of NLR proteins, particularly their variable N-terminal domains, forms the basis for their classification into distinct subfamilies: CNL (Coiled-Coil NBS-LRR), TNL (Toll/Interleukin-1 Receptor NBS-LRR), and RNL (RPW8 NBS-LRR) [2] [3]. This classification system provides a critical framework for understanding the functional specialization and evolutionary trajectories of plant immune receptors. Comparative genomic analyses across a broad spectrum of plant species have revealed remarkable variation in the abundance, distribution, and domain architecture of these subfamilies, influenced by factors such as whole-genome duplication, tandem gene amplification, and pathogen-driven selection [4] [5]. This guide objectively compares the CNL, TNL, and RNL subfamilies by synthesizing experimental data on their domain composition, phylogenetic relationships, and functional characteristics, providing researchers with a structured reference for navigating the complexity of plant NLR genes.

Domain Architecture and Classification Criteria

The canonical domain structure of NLR proteins serves as the primary criterion for subfamily classification. Each subfamily is defined by a signature N-terminal domain that dictates specific signaling functions, coupled with conserved central and C-terminal domains responsible for nucleotide binding and pathogen recognition.

• CNL (Coiled-Coil NBS-LRR): Characterized by an N-terminal coiled-coil (CC) domain, this subfamily is prevalent across all vascular plants [3] [5]. The CC domain is involved in protein-protein interactions and signaling activation. The central NB-ARC (Nucleotide-Binding Adaptor Shared by APAF-1, R Proteins, and CED-4) domain contains highly conserved motifs, including the P-loop, Kinase-2, and GLPL motifs, which facilitate ATP/GTP binding and hydrolysis [3]. A key diagnostic feature in the Kinase-2 motif is the presence of an aspartic acid (D) residue [3]. The C-terminal Leucine-Rich Repeat (LRR) domain, with its characteristic LxxLxxLxx pattern (where 'x' is any amino acid), is responsible for specific effector recognition and binding, and is subject to diversifying selection [6].

• TNL (TIR NBS-LRR): Defined by an N-terminal Toll/Interleukin-1 Receptor (TIR) domain, which shares homology with animal immune receptors [6]. The TIR domain is crucial for downstream signaling and can mediate TIR-TIR interactions for oligomerization [6]. The central NB-ARC domain is structurally similar to that of CNLs but can be distinguished by a tryptophan (W) residue in the Kinase-2 motif [3]. The C-terminal LRR domain functions in pathogen recognition. A distinctive feature of many TNLs is the presence of a C-terminal extension beyond the LRR, known as the Post-LRR (PL) domain, whose function is still being elucidated but may be involved in ligand binding or intramolecular interactions [6].

• RNL (RPW8 NBS-LRR): This subfamily features an N-terminal Resistance to Powdery Mildew 8 (RPW8) domain [7] [8]. Unlike CNLs and TNLs, which often act as pathogen sensors, RNLs primarily function as "helper" NLRs, transducing immune signals downstream of sensor NLRs [2] [8]. The NB-ARC and LRR domains maintain their conserved functions. Phylogenetically, RNLs in angiosperms are subdivided into two major clades: NRG1 (N-required gene 1) and ADR1 (activated disease resistance gene 1) [8].

Table 1: Diagnostic Features of NLR Subfamilies Based on Domain Composition

Subfamily	N-Terminal Domain	Central Domain	C-Terminal Domain	Key Diagnostic Residue (Kinase-2)	Primary Function
CNL	Coiled-Coil (CC)	NB-ARC	LRR	Aspartic Acid (D) [3]	Pathogen Sensor
TNL	TIR	NB-ARC	LRR (+PL domain in some)	Tryptophan (W) [3]	Pathogen Sensor
RNL	RPW8	NB-ARC	LRR	-	Helper/ Signal Transduction

It is important to note that many genomes contain a significant number of truncated NLR variants (e.g., NL, CN, TN, N), which lack one or more canonical domains but are still phylogenetically related to the three main subfamilies [5].

Comparative Genomic Distribution Across Plant Species

Quantitative surveys of NLR genes reveal dramatic variation in subfamily abundance and distribution across the plant kingdom, reflecting lineage-specific evolutionary paths. The following table synthesizes data from recent genomic studies.

Table 2: NLR Subfamily Distribution Across Selected Plant Species

Species	Total NLRs	CNL Count (%)	TNL Count (%)	RNL Count (%)	Key References
Arabidopsis thaliana	~150 [6]	51 (CNL & RNL) [1]	~100 [6]	(Nested within 51 CNL/RNL) [1]	[1] [6]
Glycine max (Soybean)	908 (nTNL only) [3]	467 [5]	53 [5]	31 [5]	[3] [5]
Oryza sativa (Rice)	159 (CNL only) [1]	159 [1]	0 [3]	(Identified) [3]	[1] [3]
Passiflora edulis (Purple)	25 (CNL only) [1]	25 [1]	Not Reported	Not Reported	[1]
Asparagus officinalis	27 [9]	14 (CNL & RNL) [9]	13 [9]	(Nested within 14 CNL/RNL) [9]	[9]
Cucumis sativus (Cucumber)	63 [10]	(Majority in N, NL, CNL classes) [10]	(Present in TNL class) [10]	(Present in RNL class) [10]	[10]
Prunus persica (Peach)	195 (TNL only) [6]	Not Specified	195 [6]	Not Specified	[6]
Picea mariana (Conifer)	725 (Expressed) [8]	183 (CNL) [8]	379 (TNL-related) [8]	43 (RNL-related) [8]	[8]

Key Evolutionary and Functional Insights from Comparative Data

• Monocot-Dicot Divergence: A prominent pattern is the near-complete loss of TNL genes in monocots, such as rice, while they are abundant in dicots like Arabidopsis and soybean [3]. Recent synteny-based studies suggest that the genomic regions in monocots show clear correspondence to the TNL-containing regions in dicots, explaining this absence [7].
• Lineage-Specific Expansion: The RNL subfamily, while typically small in most angiosperms, has undergone significant expansion in conifers and some Rosaceae species, suggesting a potentially enhanced role in their immune systems [8].
• Impact of Domestication: Comparative analysis of wild and cultivated species often reveals a contraction in the NLR repertoire in the domesticated form. For example, wild asparagus (Asparagus setaceus) has 63 NLRs, while cultivated garden asparagus (A. officinalis) has only 27, which may contribute to higher disease susceptibility in the crop [9].

Experimental Protocols for NLR Identification and Classification

A standardized bioinformatics workflow is essential for the accurate identification and classification of NLR genes. The following protocol, compiled from multiple studies, details the key experimental and computational steps [1] [2] [3].

Genomic Sequence Retrieval and Initial Screening

• Data Source: Obtain the complete proteome and genome annotation (GFF3 file) for the target species from public databases such as Phytozome, Ensembl Plants, or NCBI [2] [5].
• HMMER Search: Perform a Hidden Markov Model (HMM) search against the proteome using the conserved NB-ARC domain profile (Pfam: PF00931) as a query. Standard parameters include an E-value cutoff of 1e-10 to 1e-4 to ensure sensitivity [2] [4] [3].
• BLAST Enhancement: Conduct a complementary BLASTp search using known NLR reference sequences from model organisms (e.g., Arabidopsis thaliana) against the target proteome to identify divergent homologs that may be missed by HMM alone [9] [3].

Domain Validation and Architecture Analysis

• Domain Scanning: Subject all candidate sequences from the previous step to rigorous domain analysis using InterProScan, NCBI's Conserved Domain Database (CDD), and Pfam to confirm the presence of the NB-ARC domain and identify associated domains (CC, TIR, RPW8, LRR) [1] [5].
• Coiled-Coil Prediction: Use specialized tools like Paircoil2 to validate the presence of CC domains, as they can be less reliably detected by standard domain databases [1].
• Motif Identification: Use the MEME suite to identify conserved motifs within the NB-ARC domain, verifying the presence of the P-loop, Kinase-2, RNBS, and GLPL motifs. The specific residue in the Kinase-2 motif (D for CNL, W for TNL) serves as a critical diagnostic marker [2] [3].

Phylogenetic Classification and Synteny Analysis

• Sequence Alignment: Extract the NB-ARC domain sequences from all validated NLRs and perform a multiple sequence alignment using tools like ClustalW or MUSCLE [2] [3].
• Tree Construction: Construct a phylogenetic tree using the Maximum Likelihood method (e.g., with IQ-TREE or MEGA) with appropriate model selection (e.g., JTT+G+I). Bootstrap analysis with 100-1000 replicates should be used to assess node support [2] [3].
• Subfamily Assignment: Classify sequences into CNL, TNL, and RNL subfamilies based on their clustering with known reference sequences and their domain architecture [7]. Microsynteny analysis can provide further evolutionary insights, especially regarding the loss or expansion of specific subfamilies [7].

The workflow below visualizes this multi-step methodology for classifying NLR genes.

The following table catalogs key bioinformatics tools, databases, and experimental reagents essential for conducting comparative genomic analyses of NLR genes, as cited in the literature.

Table 3: Essential Research Tools and Resources for NLR Gene Analysis

Tool/Resource Name	Type	Primary Function in NLR Research	Example Use Case
Pfam [1] [2]	Database	Profile HMMs for conserved domains (e.g., NB-ARC: PF00931)	Initial identification of NLR candidates.
InterProScan [1] [5]	Software Suite	Integrated protein signature recognition	Comprehensive domain architecture analysis.
MEME Suite [2] [3]	Software	Discovery of conserved motifs in protein sequences	Identifying P-loop, Kinase-2, GLPL motifs in NB-ARC.
OrthoFinder [4]	Software	Inference of orthogroups across multiple species	Determining evolutionary relationships of NLRs across species.
IQ-TREE / MEGA [2] [9]	Software	Phylogenetic analysis using maximum likelihood	Reconstructing evolutionary history and classifying subfamilies.
PRGdb [9] [5]	Database	Curated repository of known plant R genes	Reference data for validation and comparison.
PlantCARE [9]	Database	Catalog of cis-acting regulatory elements	Analyzing promoter regions of NLR genes for stress-responsive elements.
Virus-Induced Gene Silencing (VIGS) [4]	Experimental Method	Functional validation of candidate NLR genes through transcript knockdown.	Demonstrating the role of GaNBS (OG2) in cotton leaf curl virus resistance [4].

The classification of NLR genes into CNL, TNL, and RNL subfamilies based on domain composition provides an indispensable framework for deciphering the complex landscape of plant immunity. Comparative genomics has uncovered profound diversity in the repertoire and architecture of these subfamilies across plant lineages, shaped by dynamic evolutionary processes including gene duplication, contraction, and domain fusion. The standardized experimental protocols and research tools outlined in this guide offer a roadmap for the systematic identification and functional characterization of NLR genes. As genomic data continue to accumulate, this architectural classification system will remain fundamental for discovering novel resistance genes, understanding plant-pathogen co-evolution, and ultimately engineering crops with enhanced and durable disease resistance.

Nucleotide-binding site (NBS) genes constitute the largest family of plant disease resistance (R) genes, encoding proteins that play a vital role in effector-triggered immunity against diverse pathogens [11] [1]. These genes are characterized by the presence of a conserved NBS domain, often accompanied by C-terminal leucine-rich repeats (LRRs) and variable N-terminal domains that define their classification into major subfamilies: TIR-NBS-LRR (TNL), CC-NBS-LRR (CNL), and RPW8-NBS-LRR (RNL) [11] [4]. The genomic distribution of NBS-encoding genes is not random; they frequently exhibit clustering patterns on chromosomes and are often arranged in tandem arrays, which has significant implications for their evolution and functional diversification [11] [12].

Research across numerous plant species has revealed that NBS genes are distributed unevenly across chromosomes, with a strong tendency to cluster at chromosome ends (telomeric regions) [11]. This clustering facilitates rapid evolution through mechanisms such as tandem duplication and unequal crossing over, enabling plants to generate novel resistance specificities to counter evolving pathogens [13] [12]. The study of these distribution patterns provides crucial insights into the evolutionary dynamics of plant immune systems and offers valuable resources for breeding disease-resistant cultivars through marker-assisted selection [13] [9].

Comparative Genomic Distribution of NBS Genes Across Plant Species

Chromosomal Distribution and Clustering Patterns

Table 1: Genomic Distribution of NBS Genes Across Plant Species

Plant Species	Total NBS Genes	Chromosomal Distribution	Clustered Genes	Singleton Genes	Primary Duplication Mechanism
Akebia trifoliata	73	Uneven, mostly chromosome ends	41 (56.2%)	23 (31.5%)	Tandem (33) and dispersed (29) duplications [11]
Gossypium hirsutum (TM-1)	588	Nonrandom and uneven	Tend to form clusters	Information missing	Asymmetric evolution from progenitors [12]
Gossypium barbadense	682	Nonrandom and uneven	Tend to form clusters	Information missing	Asymmetric evolution from progenitors [12]
Asparagus officinalis	27	Clustering patterns	Information missing	Information missing	Contraction during domestication [9]
Asparagus setaceus (wild)	63	Clustering patterns	Information missing	Information missing	Information missing [9]
Brassica oleracea	157	Information missing	Information missing	Information missing	Tandem duplication after whole genome triplication [14]

The distribution of NBS genes across plant genomes consistently demonstrates non-random patterns, with significant variations in gene numbers between species. In Akebia trifoliata, among 64 mapped NBS candidates, most were assigned to chromosome ends, with 41 (56.2%) located in clusters and 23 (31.5%) as singletons [11]. This telomeric preference is significant as these regions experience higher recombination rates, potentially accelerating the generation of novel resistance specificities.

Similar clustering patterns are observed in cotton species, where NBS-encoding genes display nonrandom and uneven distribution across chromosomes with a tendency to form clusters [12]. The wild asparagus species Asparagus setaceus possesses 63 NLR genes, which contracted to 47 in A. kiusianus and further reduced to just 27 in the domesticated A. officinalis, demonstrating how domestication has impacted NBS gene repertoire [9]. This contraction in cultivated species suggests artificial selection may have inadvertently reduced disease resistance capacity while selecting for other agronomic traits.

Subfamily Distribution and Architectural Diversity

Table 2: NBS Gene Subfamily Distribution Across Species

Plant Species	CNL	TNL	RNL	Other/Partial	Notable Features
Akebia trifoliata	50 (68.5%)	19 (26.0%)	4 (5.5%)	0	CNLs have fewer exons than TNLs [11]
Passiflora edulis (purple)	25	Not reported	Not reported	Not reported	Present in 3 out of 4 phylogenetic groups [1]
Gossypium arboreum	32.52% (CNL) 17.89% (CN)	3.66% (TNL) 1.63% (TN)	1.22% (RNL) 0.41% (RN)	23.98% (N) 19.51% (NL)	Higher CN/CNL, lower TNL compared to G. raimondii [12]
Gossypium raimondii	29.32% (CNL) 10.68% (CN)	25.48% (TNL) 3.83% (TN)	1.91% (RNL) 0.82% (RN)	16.99% (N) 10.96% (NL)	Higher TNL percentage (7x G. arboreum) [12]

The distribution of NBS gene subfamilies varies significantly between plant species, reflecting their distinct evolutionary paths and adaptation to different pathogen pressures. In Akebia trifoliata, the CNL subfamily dominates (68.5%), followed by TNL (26.0%) and RNL (5.5%) [11]. This pattern contrasts with cotton species, where asymmetric evolution of NBS-encoding genes is observed - Gossypium arboreum and G. hirsutum possess higher proportions of CN, CNL, and N genes, while G. raimondii and G. barbadense contain significantly more TNL genes [12].

The most striking difference between cotton species occurs in TNL type genes, with G. raimondii and G. barbadense containing approximately seven times the proportion of TNL genes compared to G. arboreum and G. hirsutum [12]. This differential distribution has functional implications, as TNL genes may play a significant role in disease resistance to Verticillium wilt in G. raimondii and G. barbadense, which are notably more resistant to this pathogen than their counterparts [12].

Methodologies for NBS Gene Identification and Analysis

Genomic Identification Pipelines

The Scientist's Toolkit: Key Research Reagents and Computational Tools for NBS Gene Analysis

Tool/Reagent Category	Specific Tools/Databases	Function in NBS Gene Research
Domain Identification	HMMER, Pfam, InterProScan, CDD, SMART	Identification of conserved NBS and associated domains (TIR, CC, LRR, RPW8) using profile hidden Markov models and domain databases [11] [9] [14]
Sequence Analysis	BLAST+, MEME Suite, CLUSTAL, MAFFT	Sequence similarity searches, motif discovery, and multiple sequence alignment [11] [4] [14]
Gene Prediction	Fgenesh++, Seqping/MAKER2, AUGUSTUS, SNAP	Ab initio and evidence-based gene prediction integrating transcriptomic and homologous protein evidence [15]
Genomic Databases	NCBI, Phytozome, BRAD, Bolbase, Plaza	Access to genomic sequences, annotations, and comparative genomics resources [4] [14]
Phylogenetic Analysis	OrthoFinder, MEGA, FastTree, DendroBLAST	Orthogroup inference, phylogenetic tree construction, and evolutionary analysis [4] [9]
Duplication Analysis	MCScanX, BEDTools, custom scripts	Identification of tandem and segmental duplications, synteny analysis [1] [9]

The accurate identification and annotation of NBS-encoding genes requires integrated computational approaches. Most studies employ a combination of Hidden Markov Model (HMM) searches and BLAST-based methods to identify candidate NBS genes [9] [14]. The standard pipeline begins with HMM searches using the conserved NB-ARC domain (PF00931) from the Pfam database as a query, typically with trusted cutoff values (e-value ≤ 1e-5 to 1e-10) [11] [14]. This is supplemented with BLAST searches against reference NLR protein sequences from model plants like Arabidopsis thaliana and Oryza sativa [9].

For domain architecture classification, identified candidates are analyzed using multiple tools including InterProScan, NCBI's Conserved Domain Database (CDD), and pairwisecoil2 or Marcoil for coiled-coil domain prediction [11] [14]. This multi-step verification ensures comprehensive identification of both typical and atypical NBS-encoding genes. High-quality gene predictions often integrate evidence from transcriptome data and homologous proteins to improve accuracy, as demonstrated in oil palm genome annotation where Fgenesh++ and Seqping pipelines were combined [15].

NBS Gene Identification and Analysis Workflow

Experimental Validation Approaches

Beyond computational identification, experimental validation is crucial for confirming NBS gene predictions and understanding their functionality. NBS profiling methods, which utilize PCR amplification with primers targeting conserved NBS motifs (P-loop, Kinase-2, and GLPL), enable experimental capture of NBS domains from genomic DNA [13]. This approach was successfully applied in potato, where just 16 amplification primers were used to generate NBS tags from 91 genomes, covering nearly all NBS domains [13].

Expression analysis through transcriptomics provides functional insights into NBS gene regulation. Studies typically examine expression patterns across different tissues, developmental stages, and under various stress conditions [11] [4]. For instance, in Akebia trifoliata, NBS genes were generally expressed at low levels, with a few showing relatively high expression during later development in rind tissues [11]. Functional validation often employs virus-induced gene silencing (VIGS), as demonstrated in cotton where silencing of GaNBS (OG2) revealed its putative role in virus tittering [4].

Evolutionary Mechanisms Shaping NBS Gene Distribution

Duplication Mechanisms and Selection Pressures

The expansion and diversification of NBS gene families are primarily driven by various duplication mechanisms, with tandem and dispersed duplications recognized as the main forces responsible for NBS gene proliferation [11]. In Akebia trifoliata, tandem duplications produced 33 genes while dispersed duplications generated 29 genes [11]. Similarly, in passion fruit, CNL genes expanded through both segmental (17 gene pairs) and tandem duplications (17 gene pairs) [1].

The evolutionary history of plant genomes significantly influences NBS gene distribution. In Brassica species, whole genome triplication (WGT) of the Brassica ancestor followed by extensive gene loss shaped the current NBS gene repertoire [14]. After WGT, NBS-encoding homologous gene pairs on triplicated regions were rapidly deleted or lost, with subsequent species-specific gene amplification occurring through tandem duplication after the divergence of B. rapa and B. oleracea [14].

Selection pressure analyses reveal that NBS genes typically undergo strong purifying selection, which maintains conserved functional domains while allowing variation in pathogen recognition regions [1] [14]. Evolutionary studies of CNL-type NBS-encoding orthologous gene pairs between Brassica species and Arabidopsis indicated that orthologous genes in B. rapa have undergone stronger negative selection than those in B. oleracea [14].

Evolutionary Mechanisms Shaping NBS Gene Distribution

Impact of Domestication on NBS Gene Repertoires

Comparative analyses between wild and cultivated species provide compelling evidence for the impact of domestication on NBS gene repertoires. In asparagus, a marked contraction of NLR genes occurred from wild species to the domesticated A. officinalis, with gene counts reduced from 63 in A. setaceus to 47 in A. kiusianus and only 27 in A. officinalis [9]. This reduction in NBS gene diversity during domestication likely contributes to the increased disease susceptibility observed in cultivated varieties.

Orthologous gene analysis between A. setaceus and A. officinalis identified only 16 conserved NLR gene pairs, representing the NLR genes preserved during the domestication process of A. officinalis [9]. Notably, the majority of preserved NLR genes in A. officinalis demonstrated either unchanged or downregulated expression following fungal challenge, indicating potential functional impairment in disease resistance mechanisms as a consequence of artificial selection favoring yield and quality traits over disease resistance [9].

The genomic distribution patterns of NBS genes, characterized by chromosomal clustering and tandem arrangements, reflect evolutionary adaptations to relentless pathogen pressure. These distribution patterns are conserved across plant species yet exhibit species-specific variations in subfamily composition and cluster organization. The tendency for NBS genes to form clusters, particularly in telomeric regions, facilitates rapid evolution through mechanisms like tandem duplication and unequal crossing over, enabling plants to continuously generate novel resistance specificities.

Understanding these distribution patterns has significant practical implications for crop improvement. Molecular markers developed from NBS gene clusters can enable marker-assisted selection for disease resistance breeding [13]. The comparative genomics approaches outlined in this review facilitate identification of key resistance genes in wild relatives that can be introgressed into cultivated varieties. Furthermore, knowledge of NBS gene evolution and distribution informs development of durable resistance strategies that can counter pathogen evolution and mitigate yield losses in agricultural production systems.

Future research directions should include more comprehensive comparative analyses across broader phylogenetic ranges, integration of pan-genome approaches to capture species-level diversity, and functional characterization of clustered NBS genes to elucidate their roles in pathogen recognition and defense signaling. Such advances will continue to enhance our understanding of plant immunity and contribute to the development of sustainable crop protection strategies.

The study of genomic evolutionary dynamics, specifically the expansion and contraction of gene families, provides a critical window into understanding how plants adapt to environmental stresses, evolve developmental complexity, and generate biodiversity. Among the most dynamic components of plant genomes are Nucleotide-Binding Site (NBS) domain genes, which constitute a major class of disease resistance (R) genes that plants employ in pathogen defense mechanisms [4]. Recent comparative genomic analyses across diverse plant lineages have revealed that these genes undergo remarkably dynamic evolutionary changes, including rapid expansion, contraction, and functional diversification, often driven by selective pressures from evolving pathogen populations [16] [4]. The investigation of these patterns provides not only fundamental insights into plant evolutionary biology but also practical avenues for crop improvement through the identification of novel resistance elements.

This guide objectively compares the evolutionary dynamics of NBS domain genes across multiple plant species, synthesizing data from recent large-scale genomic studies to elucidate patterns of gene family expansion and contraction. We present comprehensive comparative data, detailed experimental methodologies for analyzing these evolutionary trajectories, and visualizations of the underlying biological processes, providing researchers with a framework for investigating genomic evolution in plant systems.

Comparative Analysis of NBS Gene Family Dynamics Across Plant Lineages

Evolutionary Patterns and Species-Specific Expansions

Table 1: Evolutionary Patterns of NBS Domain Genes Across Plant Species

Plant Species	Genome Characteristics	NBS Gene Count	Expansion Mechanisms	Evolutionary Features
Brassica carinata (zd-1)	Allotetraploid (BBCC); ~1.1 Gbp	2,570 RGAs (2020 TM-LRR, 550 NBS-LRR) [17]	Intergenomic/intragenomic duplications (65.2% of RGAs) [17]	Subgenome dominance; Extensive RGA expansion compared to progenitors [17]
Barley (Hordeum vulgare 'Morex V3')	Diploid cereal crop	214 significantly expanded orthogroups [18]	Tandem and segmental duplications [18]	Evolve more rapidly with lower negative selection; lower GC content [18]
Cowpea (Vigna unguiculata 'CPD103')	Diploid legume; 641 Mbp	2,188 R-genes (29 classes) [19]	Dispersed and tandem duplication under purifying selection [19]	Kinases (KIN) and transmembrane proteins (RLKs/RLPs) prominent [19]
Passion fruit (Passiflora edulis Sims.)	Diploid fruiting crop	25 CNL genes [20]	Segmental (17 pairs) and tandem (17 pairs) duplications [20]	Strong purifying selection; clustered on chromosome 3 [20]
Angiosperms (304 species)	Diverse ploidy levels	>90,000 NLR genes (18,707 TNL, 70,737 CNL, 1,847 RNL) [4]	Whole genome duplication and small-scale duplications [4]	Massive expansion in flowering plants compared to non-flowering plants [4]
Bryophytes (e.g., Physcomitrella patens)	Early land plants	~25 NLR genes [4]	Limited duplication events	Compact NLR repertoires representing ancestral states [4]

The comparative data reveal striking differences in NBS gene family sizes and architectures across plant lineages. Flowering plants exhibit substantial expansions in their NBS gene repertoires compared to non-flowering plants, with angiosperms collectively encoding over 90,000 NLR genes across 304 species surveyed [4]. This represents a dramatic increase from the approximately 25 NLR genes found in bryophytes like Physcomitrella patens, suggesting that the evolutionary transition to flowering plants was accompanied by massive diversification of disease resistance genes [4].

Polyploid species demonstrate particularly complex evolutionary patterns, as evidenced by Brassica carinata, where 65.2% of resistance gene analogs (RGAs) show evidence of gene duplication events, with contrasting patterns between subgenomes indicating subgenome dominance [17]. This phenomenon of subgenome dominance in allopolyploids appears to be a shared characteristic across Brassica species and significantly influences how gene families expand and contract following genome duplication events.

Molecular Mechanisms Driving Gene Family Dynamics

Table 2: Molecular Mechanisms of Gene Family Expansion and Contraction

Mechanism	Molecular Process	Impact on Gene Family	Examples
Whole Genome Duplication (WGD)	Doubling of entire genome	Creates numerous paralogs; provides raw material for neofunctionalization [18]	Found in all angiosperms; brassica species [17] [18]
Tandem Duplication	Localized duplication of chromosomal segments	Creates gene clusters; rapid expansion of specific gene families [4]	NBS-LRR genes in passion fruit (17 tandem pairs) [20]
Segmental Duplication	Duplication of large chromosomal regions	Distributed gene duplicates; conservation of gene order [4]	Passion fruit (17 segmental pairs) [20]
Transposable Element-Mediated Duplication	TE activity facilitates gene duplication	Rapid emergence of novel gene arrangements [21]	Association with 30-40% of de novo genes in rice/maize [21]
Gene Conversion	Non-reciprocal transfer of genetic information	Homogenization of gene families; concerted evolution [22]	Observed in Asteraceae R-genes [22]
De Novo Gene Origination	Emergence from non-coding DNA	Totally novel genes without precursors [21]	OsDR10 in rice, AtQQS in Arabidopsis [21]

The evolutionary trajectories of plant gene families are shaped by multiple molecular mechanisms. Whole-genome duplication (WGD) events provide the primary substrate for gene family expansion in flowering plants, with numerous documented WGD events in species including rice, maize, and cotton [18]. These duplicated genomes subsequently undergo a process of fractionation and diploidization, where many duplicated genes are lost while others are retained through processes of neofunctionalization (where one copy acquires a new function), subfunctionalization (where ancestral functions are partitioned between duplicates), or dosage advantage (where increased gene copy number provides selective benefit) [18].

Recently, the role of de novo gene origination from previously non-coding DNA has gained recognition as a significant contributor to genetic novelty. Plant genomes are particularly conducive to this process due to their expansive non-coding regions and high transposable element content, which provides rich substrate for novel gene birth [21]. These de novo genes typically encode shorter proteins with high intrinsic disorder content, lacking recognizable conserved domains, which may facilitate rapid functional exploration [21].

Experimental Approaches for Analyzing Gene Family Evolution

Genomic Identification and Annotation of NBS Domain Genes

The comprehensive identification and classification of NBS domain genes requires integrated bioinformatics approaches. The standard workflow begins with whole-genome sequencing using either Illumina short-read or Nanopore long-read technologies, or often a hybrid approach for optimal assembly, as demonstrated in cowpea [19]. Following genome assembly and repeat masking, NBS domain genes are typically identified using Hidden Markov Model (HMM) searches against the Pfam database, specifically targeting the NB-ARC domain (PF00931) [18] [4].

OrthoFinder is commonly employed for orthogroup clustering across multiple species, enabling the differentiation between orthologs (genes in different species that evolved from a common ancestral gene) and paralogs (genes related by duplication within a genome) [18]. For the specific identification of CNL (CC-NBS-LRR) genes, as performed in passion fruit, a combination of BLASTp searches using known CNL proteins from reference species like Arabidopsis thaliana coupled with domain verification through Pfam, CDD, and InterProScan provides robust identification [20]. This multi-step verification ensures comprehensive detection while minimizing false positives.

Evolutionary Analysis and Selection Pressure Assessment

To elucidate evolutionary relationships and selection pressures, researchers employ phylogenetic reconstruction and evolutionary rate calculations. Multiple sequence alignment using tools like MAFFT or Clustal provides the basis for phylogenetic tree construction, typically performed with maximum likelihood algorithms implemented in FastTreeMP or similar programs [4]. These phylogenetic analyses reveal deep evolutionary relationships and can identify lineage-specific expansion events.

The assessment of selection pressures represents a crucial component of evolutionary analysis. The non-synonymous (Ka) to synonymous (Ks) substitution rate ratio (Ka/Ks) serves as a key metric for identifying evolutionary forces acting on gene families [18]. Ka/Ks ratios significantly less than 1 indicate purifying selection, ratios approximately equal to 1 suggest neutral evolution, and ratios greater than 1 provide evidence for positive selection [18]. In barley, for example, expanded genes were found to evolve more rapidly and experience lower negative selection pressure compared to non-expanded genes [18].

Figure 1: Experimental workflow for analyzing gene family evolution, showing the progression from genome assembly through identification, evolutionary analysis, and functional validation.

Functional Validation of Expanded Gene Families

Following computational identification and evolutionary analysis, functional validation provides critical evidence for the biological roles of expanded gene families. Expression profiling using RNA-seq data under various stress conditions or across different tissues helps associate candidate genes with specific biological processes [4] [20]. For example, in passion fruit, PeCNL3, PeCNL13, and PeCNL14 were identified as differentially expressed under Cucumber mosaic virus infection and cold stress [20].

For direct functional testing, virus-induced gene silencing (VIGS) has proven effective in validating disease resistance genes. In cotton, silencing of GaNBS (OG2) demonstrated its putative role in virus tittering, confirming its function in disease resistance [4]. Additionally, emerging machine learning approaches are being employed to identify multi-stress responsive genes, as demonstrated in passion fruit where a Random Forest model successfully validated three CNL genes as multi-stress responsive [20].

Table 3: Essential Research Reagents and Computational Tools for Evolutionary Genomics

Category	Specific Tools/Reagents	Application	Key Features
Sequencing Technologies	Illumina HiSeq X Ten, Oxford Nanopore GridION X5 [19]	Whole genome sequencing	Short-read vs. long-read complementarity; hybrid assembly approaches
Genome Assembly	MaSuRCA v3.4.2 [19]	Hybrid genome assembly	Integrates both short and long reads for optimal contiguity
Gene Identification	HMMER, PfamScan, OrthoFinder v2.5.4 [18] [4]	Domain identification and orthogroup clustering	Hidden Markov Models for domain detection; orthology assignment
Evolutionary Analysis	MAFFT, FastTreeMP, PAML CODEML [18] [4]	Phylogenetics and selection pressure	Multiple sequence alignment; Ka/Ks calculation
Expression Analysis	RNA-seq, qPCR [23] [20]	Expression profiling	Tissue-specific and stress-responsive expression patterns
Functional Validation	VIGS, CRISPR/Cas9 [4] [21]	Gene function determination	Transient silencing; targeted mutagenesis
Data Resources	NCBI, Phytozome, Plaza, Ensembl Plants [4] [20]	Genomic data repositories	Curated genome assemblies and annotations

This toolkit represents the essential resources required for comprehensive evolutionary genomics studies of plant gene families. The combination of sequencing technologies provides the fundamental data, while bioinformatic tools enable the identification and evolutionary analysis of gene families of interest. Functional validation techniques then bridge computational predictions with biological reality, creating a闭环 research pipeline from gene identification to functional characterization.

The comparative analysis of expansion and contraction patterns across plant lineages reveals NBS domain genes as exceptionally dynamic components of plant genomes, characterized by repeated cycles of duplication, functional diversification, and occasional loss. These evolutionary processes create genetically diverse repertoires of disease resistance genes that enable plants to adapt to evolving pathogen pressures. The experimental frameworks outlined herein provide researchers with robust methodologies for investigating these evolutionary trajectories, while the visualization approaches and reagent toolkit offer practical resources for implementing these analyses. As genomic technologies continue to advance, particularly in long-read sequencing and genome editing, our ability to decipher the complex evolutionary dynamics of plant gene families will continue to deepen, offering new insights for both basic plant evolutionary biology and applied crop improvement strategies.

The nucleotide-binding site-leucine-rich repeat (NBS-LRR) gene family constitutes a critical component of the plant immune system, encoding intracellular receptors that recognize pathogen effectors and initiate effector-triggered immunity [24] [25]. The size and composition of this gene family exhibit remarkable variation across the plant kingdom, reflecting diverse evolutionary paths and adaptation strategies. This guide provides a comparative analysis of NBS family size variation from early land plants like mosses to advanced angiosperms, synthesizing quantitative data and methodological approaches to elucidate lineage-specific adaptations in plant immunity.

NBS-LRR genes represent one of the largest and most variable gene families in plants, with dramatic expansions and contractions occurring throughout plant evolution [4] [9]. The proliferation of these genes is primarily driven by various duplication mechanisms, including whole-genome duplication (WGD) and small-scale duplication events, which provide raw genetic material for innovation in pathogen recognition [26] [27]. Understanding the patterns of NBS family size variation across different plant lineages offers insights into the evolutionary mechanisms shaping plant-pathogen interactions and informs strategies for crop improvement through manipulation of resistance genes.

Comparative Genomic Analysis of NBS Family Size Across Plant Lineages

Quantitative Variation in NBS-LRR Genes

Table 1: NBS-LRR Gene Family Size Variation Across Plant Species

Plant Species	Lineage Group	Total NBS Genes	CNL/Non-TNL	TNL	RNL	Other/Variants	Primary Expansion Mechanism
Physcomitrella patens (moss)	Bryophyte	~25	Not specified	Not specified	Not specified	Not specified	Not specified
Selaginella moellendorffii (spikemoss)	Lycophyte	~2	Not specified	Not specified	Not specified	Not specified	Not specified
Asparagus setaceus (wild)	Monocot	63	Not specified	Not specified	Not specified	Not specified	Natural selection
Asparagus kiusianus (wild)	Monocot	47	Not specified	Not specified	Not specified	Not specified	Natural selection
Asparagus officinalis (domesticated)	Monocot	27	Not specified	Not specified	Not specified	Not specified	Contraction during domestication
Nicotiana sylvestris	Eudicot	344	82 (CC-NBS) 48 (CC-NBS-LRR)	5 (TIR-NBS) 37 (TIR-NBS-LRR)	Not specified	172 (NBS-only)	Whole-genome duplication
Nicotiana tomentosiformis	Eudicot	279	65 (CC-NBS) 47 (CC-NBS-LRR)	7 (TIR-NBS) 33 (TIR-NBS-LRR)	Not specified	127 (NBS-only)	Whole-genome duplication
Nicotiana tabacum	Eudicot	603	150 (CC-NBS) 74 (CC-NBS-LRR)	9 (TIR-NBS) 64 (TIR-NBS-LRR)	Not specified	306 (NBS-only)	Allotetraploidization + WGD
Akebia trifoliata	Eudicot	73	Not specified	Not specified	Not specified	Not specified	Not specified
Vitis vinifera	Eudicot	352	Not specified	Not specified	Not specified	Not specified	Not specified
Triticum aestivum (bread wheat)	Monocot	1,500-2,151	Not specified	Not specified	Not specified	Not specified	Polyploidization

The data reveal several key patterns in NBS family evolution. Bryophytes and lycophytes maintain relatively small NBS repertoires (approximately 25 and 2 genes, respectively), indicating that substantial gene expansion occurred primarily in flowering plants [4]. Among angiosperms, significant variation exists, with domesticated species like Asparagus officinalis showing marked contraction (27 genes) compared to its wild relatives (47-63 genes), suggesting that artificial selection for agronomic traits may reduce immune gene diversity [9]. Allotetraploid species such as Nicotiana tabacum demonstrate the profound impact of whole-genome duplication, possessing approximately twice the NBS gene count (603 genes) of its diploid progenitors [28].

Lineage-Specific Trends in NBS Family Composition

Different plant lineages show distinct patterns of NBS gene expansion and contraction. In Solanaceae species, NBS-LRR genes are predominantly of the CNL type, with TNLs representing a smaller proportion. A study of nine Solanaceae species identified 819 NBS-LRR genes, comprising 583 CNL (71.2%), 182 TNL (22.2%), and 54 RNL (6.6%) genes [25]. This distribution contrasts with patterns in other plant families, suggesting lineage-specific selection pressures.

Notably, complete loss of TNL genes has occurred in some lineages, including the Poaceae family and the dicot Mimulus guttatus [24]. This pattern indicates that different plant lineages have evolved distinct strategies for pathogen recognition, with some emphasizing CNL-type genes while largely abandoning TNL-type genes.

Methodological Framework for NBS Gene Identification and Analysis

Standardized Bioinformatics Workflow

Table 2: Experimental Protocols for NBS Gene Family Analysis

Methodological Step	Standard Tools/Approaches	Key Parameters	Application in NBS Studies
Gene Identification	HMMER search with PF00931 (NB-ARC domain)	E-value cutoff: 1e-5 to 1e-10; domain completeness verification	Initial screening of genomic sequences for NBS domain candidates [28] [9]
Domain Architecture Analysis	InterProScan, NCBI CDD, Pfam database	Domain E-value threshold: 1e-5; manual curation of domain boundaries	Classification into CNL, TNL, RNL, and truncated variants [4] [9]
Phylogenetic Analysis	MUSCLE/Clustal Omega for alignment; MEGA for tree construction	JTT model; 1000 bootstrap replicates; maximum likelihood method	Evolutionary relationships within and between species [28] [9]
Duplication Pattern Analysis	MCScanX, BLASTP all-vs-all search	E-value: 1e-5; collinearity detection; synteny analysis	Identification of WGD, tandem, proximal, and dispersed duplications [25] [29] [28]
Selection Pressure Analysis	KaKs_Calculator with Nei-Gojobori method	Ka/Ks ratio calculation: >1 positive selection, <1 purifying selection, =1 neutral evolution	Detection of evolutionary forces acting on NBS genes [28]
Expression Analysis	RNA-seq alignment (HISAT2), quantification (Cufflinks)	FPKM normalization; differential expression (Cuffdiff)	Expression patterns under biotic stress and in different tissues [4] [28]

The consistent application of these methodologies across studies enables comparative analyses and meta-analyses of NBS gene families across diverse plant species. The integration of multiple bioinformatics tools creates a robust pipeline for comprehensive NBS gene identification and characterization.

Visualization of NBS Gene Analysis Workflow

NBS Gene Analysis Workflow

Mechanisms Driving NBS Family Expansion and Contraction

Gene Duplication Modalities

The expansion of NBS gene families primarily occurs through various duplication mechanisms, each contributing differently to gene family evolution:

• Whole-Genome Duplication (WGD): WGD events simultaneously duplicate all genes in the genome, providing substantial raw material for NBS family expansion. In Solanaceae species, WGD has played a particularly important role in NBS-LRR gene expansion [25]. Allotetraploid species like Nicotiana tabacum show approximately double the NBS gene count compared to its diploid progenitors, demonstrating the significant impact of WGD [28].

• Tandem Duplication (TD): Tandem duplication occurs through unequal crossing over and generates clusters of similar genes in close chromosomal proximity. This mechanism is prevalent in plant genomes and contributes significantly to the rapid expansion of NBS genes in response to pathogen pressure [26]. Tandem duplicates often undergo rapid functional divergence, allowing for the generation of new pathogen recognition specificities [26] [29].

• Proximal Duplication (PD): Proximal duplication involves genes located close together on chromosomes but separated by a few genes. These may represent ancient tandem duplicates that have been disrupted by the insertion of other genes over evolutionary time [29].

• Transposed Duplication (TRD): Transposed duplication involves the relocation of gene copies to new chromosomal positions through DNA-based or RNA-based (retrotransposition) mechanisms. Retrotransposed duplicates often show higher expression and regulatory divergence compared to other duplication types [29].

• Dispersed Duplication (DSD): Dispersed duplication generates duplicated genes that are scattered throughout the genome without clear patterns of collinearity. The mechanisms underlying dispersed duplication remain less understood but contribute significantly to NBS family diversity [26].

Evolutionary Fate of Duplicated NBS Genes

Following duplication, NBS genes undergo various evolutionary processes that determine their retention or loss:

• Purifying Selection: Most duplicated NBS genes are under purifying selection, which removes deleterious mutations while preserving gene function [26]. This is evidenced by Ka/Ks ratios less than 1 in studies of duplicated genes in Aurantioideae [26].

• Positive Selection: Specific codons in NBS genes, particularly in the LRR domain, often experience positive selection that drives functional diversification and enables recognition of evolving pathogen effectors [30].

• Nonfunctionalization: Many duplicated NBS genes accumulate deleterious mutations and become pseudogenes, eventually being lost from the genome through deletion or sequence degeneration.

• Neofunctionalization: Some duplicates acquire new functions through accumulation of mutations, potentially generating novel pathogen recognition specificities [27] [29].

• Subfunctionalization: Duplicates may partition ancestral functions between them, with each copy specializing in certain aspects of the original gene's function [29].

Table 3: Research Reagent Solutions for NBS Gene Studies

Reagent/Resource	Function	Example Applications	Key Features
HMMER Suite	Hidden Markov Model-based sequence search	Identification of NBS domains using PF00931 profile	Sensitive detection of divergent NBS domains; customizable thresholds [28] [9]
MCScanX	Detection of gene duplication patterns	Identification of WGD, tandem, and proximal duplications	Collinearity analysis; visualization of syntenic blocks [25] [29] [28]
PFAM Database	Protein family and domain annotation	Classification of NBS, TIR, CC, LRR domains	Curated domain models; functional annotations [4] [9]
OrthoFinder	Orthogroup inference and comparative genomics	Identification of orthologous NBS genes across species	Accurate orthogroup prediction; phylogenetic species tree reconstruction [4]
KaKs_Calculator	Calculation of selection pressures	Ka/Ks analysis for detecting positive selection	Multiple evolutionary models; statistical reliability [28]
PlantCARE	Identification of cis-regulatory elements	Analysis of promoter regions of NBS genes	Database of plant cis-elements; prediction of regulatory motifs [9]
PRGdb	Plant Resistance Gene database	Classification and annotation of NBS-LRR genes	Curated R-gene database; functional classifications [24] [9]

These resources form the foundation of contemporary comparative genomics studies of NBS gene families, enabling researchers to identify, classify, and analyze evolutionary patterns across plant species.

Visualization of NBS Domain Architecture and Classification

NBS Protein Domain Architecture and Classification

The comparative analysis of NBS gene family size across plant lineages reveals a complex evolutionary history shaped by diverse mechanisms. Bryophytes maintain modest NBS repertoires, while angiosperms demonstrate dramatic expansions through both whole-genome and small-scale duplication events [31] [4]. Lineage-specific patterns, such as the complete loss of TNL genes in Poaceae and the contraction of NBS families during domestication in Asparagus officinalis, highlight the dynamic nature of plant immune gene evolution [24] [9].

The variation in NBS family size and composition reflects different evolutionary strategies for pathogen recognition, with some lineages emphasizing diversity through gene duplication while others may optimize for efficiency with smaller, more versatile repertoires. Understanding these lineage-specific adaptations provides fundamental insights into plant immunity and offers potential strategies for engineering disease resistance in crop species through manipulation of NBS gene content and diversity.

Future research directions should include more comprehensive sampling across plant lineages, functional characterization of NBS genes in non-model species, and investigation of the relationship between NBS repertoire size and ecological factors such as pathogen pressure and life history traits. Such studies will further illuminate the evolutionary forces shaping this critical component of the plant immune system.

Nucleotide-binding leucine-rich repeat receptors (NLRs) represent the largest and most variable class of intracellular immune receptors in plants, serving as critical components of the effector-triggered immunity (ETI) system [9] [32]. These genes exhibit exceptional diversity both within and across plant species, with their sequences and genomic distributions bearing the imprints of past evolutionary pressures, including plant-pathogen co-evolution and major speciation events [33] [32]. The comparative analysis of NLR genes across related species provides a powerful framework for reconstructing phylogenetic relationships and tracing the evolutionary history of plant lineages. Recent advances in genomic sequencing and bioinformatic tools have enabled researchers to comprehensively identify NLR repertoires (NLRomes) across multiple species, revealing complex patterns of gene expansion, contraction, and diversification that often correlate with significant evolutionary transitions [34] [35]. This guide systematically compares the experimental approaches, computational tools, and analytical frameworks currently employed in NLR-based phylogenetic reconstruction, providing researchers with practical methodologies for investigating plant evolutionary history through the lens of immune gene evolution.

Methodological Framework: Comparative Genomics of NLR Genes

Core Workflow for NLR Identification and Phylogenetic Analysis

The standard pipeline for NLR-based phylogenetic reconstruction integrates genome-wide gene identification, evolutionary analysis, and phylogenetic inference, with specialized tools available for each stage. The following diagram illustrates the core workflow:

NLR Identification and Annotation Tools Comparison

Accurate identification of NLR genes is the foundational step in phylogenetic analysis. Different tools vary in their approaches and performance characteristics:

Table 1: Comparison of NLR Identification Tools and Methods

Tool/Method	Approach	Advantages	Limitations	Best Applications
NLRSeek [34]	Genome reannotation-based pipeline	Identifies previously missed NLRs; 33.8%-127.5% more NLRs in yam species; validates expression	Computationally intensive; requires genomic sequences	Non-model species with incomplete annotations
HMMER Search [9]	Hidden Markov Models with NB-ARC domain (PF00931)	High specificity for conserved domains; standardized approach	May miss divergent or truncated NLRs	Initial screening in well-annotated genomes
BLAST-based Methods [9]	Sequence similarity to known NLR references	Fast; good for preliminary identification	Reference-dependent; may miss novel NLR lineages	Cross-species comparison with established references
Combined Approach [9]	Integrates HMMER and BLAST with manual validation	Comprehensive coverage; reduces false negatives	Labor-intensive; requires expert curation	Critical studies requiring complete NLR repertoires

Experimental Protocols for NLR Gene Family Analysis

Genome-Wide NLR Identification Protocol

The standard protocol for comprehensive NLR identification combines multiple complementary approaches [9] [34]:

• Data Acquisition: Obtain chromosomal-level genome assemblies and annotation files for target species. High-quality assemblies with high BUSCO completeness scores (>97%) are essential for comprehensive identification [9].

• Initial Candidate Identification:

  • Perform HMMER searches using the NB-ARC domain (PF00931) profile with an E-value cutoff of 1e-5
  • Conduct local BLASTp searches against reference NLR proteins from related species with E-value ≤ 1e-10
  • Extract candidate sequences using bioinformatics tools like TBtools [9]

• Domain Validation and Classification:

  • Verify domain architecture using InterProScan and NCBI's Batch CD-Search
  • Classify NLRs into subfamilies (CNL, TNL, RNL) based on N-terminal domains
  • Identify truncated variants (NL, CN, TN, RN) lacking specific domains [9]

• Manual Curation and Validation:

  • Reconcile predictions with existing annotations
  • Perform targeted genome reannotation for missed NLRs using NLRSeek pipeline [34]
  • Validate expression through transcriptomic data where available

Phylogenetic Reconstruction Methodology

The standard phylogenetic analysis protocol involves [9] [36]:

• Sequence Alignment: Perform multiple sequence alignment of NLR protein sequences using Clustal Omega or MAFFT with default parameters.

• Tree Construction: Build phylogenetic trees using maximum likelihood method (e.g., MEGA, RAxML) based on the JTT matrix-based model with 1000 bootstrap replicates.

• Evolutionary Analysis:

  • Identify orthologous gene pairs using OrthoFinder
  • Analyze evolutionary patterns (expansion/contraction) by comparing gene counts across species
  • Detect conserved NLR lineages preserved through speciation events

Comparative Genomic Analyses: Case Studies Across Plant Families

NLR Repertoire Variation Across Plant Lineages

Different plant families exhibit distinct evolutionary patterns in their NLRomes, reflecting varied evolutionary histories and selection pressures:

Table 2: NLR Repertoire Comparisons Across Plant Families

Plant Family/Species	NLR Count	Evolutionary Pattern	Key Findings	Evolutionary Drivers
Asparagus species [9]	A. setaceus: 63A. kiusianus: 47A. officinalis: 27	Contraction in cultivated species	16 conserved orthologous pairs identified; susceptibility linked to repertoire reduction	Domestication pressure favoring yield over immunity
Vicioid legumes [35]	Variable across tribes: Cicereae/Fabeae (contraction)Trifolieae (expansion)	Tribe-specific expansion/contraction	Recent expansion in Trifolieae (1-6 Mya) with higher substitution rates	Whole genome duplication followed by diploidization
Dendrobium orchids [36]	655 NBS genes across 7 species	Lineage-specific degeneration	TNL absence in monocots; degeneration on specific phylogenetic branches	NRG1/SAG101 pathway deficiency in monocots
Oleaceae family [37]	Fraxinus: ConservationOlea: Expansion	Genus-specific strategies	Fraxinus: conserved genesOlea: recent duplications and novel NLR births	Geographical adaptation; differential pathogen pressures
General range [32]	<100 to >1,000 per genome	Rapid birth-death evolution	Correlation with total gene number; exception in specific lineages (e.g., cucurbits)	Pathogen-driven selection; fitness costs of NLR maintenance

Visualization and Analysis Tools for Phylogenetic Data

Effective visualization of phylogenetic trees is essential for interpreting complex evolutionary relationships:

Table 3: Phylogenetic Tree Visualization Tools Comparison

Tool/Software	Primary Features	Visualization Capabilities	Annotation Options	Best Use Cases
ggtree [38]	R package, ggplot2 integration	Rectangular, circular, fan, unrooted layouts	Extensive annotation layers; taxonomic coloring	Publication-quality figures; complex data integration
Archaeopteryx [39]	Java-based desktop application	Standard tree layouts with rotation capability	Taxonomic metadata from databases; color by taxonomy	Interactive tree exploration; taxonomic analysis
ColorPhylo [40]	Automatic color coding method	Any tree visualization platform	Colors reflect taxonomic distances	Intuitive display of taxonomic relationships
iTOL/FigTree [38]	Web-based/desktop applications	Standard phylogenetic layouts	Pre-defined annotation functions	Quick visualization; standard phylogenetic workflows

Computational Tools and Databases

Successful NLR phylogenetic analysis requires specialized computational resources and biological materials:

Table 4: Essential Research Reagents and Resources for NLR Phylogenetics

Category	Specific Tools/Resources	Function/Purpose	Key Features
Genomic Databases	Plant GARDEN [9], Dryad Digital Repository [9], NCBI Taxonomy	Source of genomic and taxonomic data	Chromosomal-level assemblies; standardized annotations
NLR Identification	NLRSeek [34], HMMER, InterProScan [9]	Comprehensive NLR mining and annotation	Genome reannotation; domain architecture analysis
Sequence Analysis	Clustal Omega [9], MEME suite [9], PlantCARE [9]	Multiple alignment, motif discovery, cis-element analysis	Conserved motif identification; promoter element prediction
Phylogenetic Analysis	MEGA [9], OrthoFinder [9], ggtree [38]	Tree construction, orthology assessment, visualization	Maximum likelihood methods; orthogroup inference
Expression Validation	RNA-seq datasets (SRA) [37], WoLF PSORT [9]	Expression analysis; subcellular localization	Experimental validation of NLR function

Experimental Workflow Integration

The integration of computational predictions with experimental validation creates a powerful framework for evolutionary analysis. The following diagram illustrates the relationship between key analytical components and their outputs in NLR phylogenetic studies:

Discussion: Interpretation of Evolutionary Patterns in NLR Phylogenies

Key Evolutionary Patterns and Their Significance

Phylogenetic analyses of NLR genes across multiple plant families have revealed consistent evolutionary patterns that provide insights into plant evolutionary history:

Differential Expansion and Contraction - Different plant lineages exhibit distinct trajectories of NLR repertoire evolution. The significant contraction observed in domesticated asparagus (from 63 NLRs in wild A. setaceus to 27 in cultivated A. officinalis) demonstrates how artificial selection can reshape immune gene repertoires, potentially at the cost of disease susceptibility [9]. Conversely, the expansion in Trifolieae legumes illustrates how specific lineages can rapidly diversify their immune receptors in response to pathogen pressures [35].

Lineage-Specific Subfamily Dynamics - The absence of TNL genes in monocots, including orchids and grasses, represents a major evolutionary transition in plant immunity, possibly driven by the loss of downstream signaling components [36]. This pattern serves as a valuable phylogenetic marker for deep evolutionary relationships.

Conserved Orthologous Lineages - The identification of conserved NLR pairs across species, such as the 16 orthologous groups preserved between wild and cultivated asparagus, highlights immune genes maintained over evolutionary timeframes, potentially representing core components of the plant immune system [9].

Technical Considerations and Methodological Recommendations

Based on comparative analyses of current research, several recommendations emerge for NLR-based phylogenetic studies:

• Employ Complementary Identification Methods - Studies consistently identify more NLR genes using integrated approaches (e.g., NLRSeek identified 33.8%-127.5% more NLRs in yam species compared to conventional methods) [34]. The combination of HMM-based and similarity-based approaches with manual curation provides the most comprehensive NLR repertoires.

• Account for Taxonomic Sampling Biases - Evolutionary interpretations must consider the uneven taxonomic sampling and varying genome quality across species. The use of high-quality chromosomal-level assemblies improves comparative analyses.

• Integrate Expression Data - Phylogenetic patterns gain functional context when correlated with expression data. In olive, partially structured NLR genes show significant expression despite incomplete domains, suggesting potential functional importance [37].

• Consider Evolutionary Time Scales - Different evolutionary processes operate at different time scales. Recent duplications (1-6 Mya in Trifolieae) [35] versus ancient whole genome duplications (~35 Mya in Fraxinus) [37] leave distinct signatures in NLR phylogenies that require different interpretive frameworks.

This comparative guide provides researchers with the methodological foundation and analytical frameworks necessary to reconstruct plant evolutionary history through NLR gene phylogenies, contributing to a deeper understanding of how immune gene evolution has shaped plant diversity.

From Genomes to Annotations: Computational Pipelines for NBS Gene Identification

In the field of plant comparative genomics, particularly in the study of nucleotide-binding site (NBS) domain genes, bioinformatics tools form the cornerstone of discovery. NBS domain genes represent one of the largest superfamilies of plant resistance genes, playing crucial roles in pathogen recognition and defense activation [4]. The exponential growth of genomic data from diverse plant species has created an pressing need for robust bioinformatics workflows that can identify and characterize these important genetic elements across taxa. Among the most critical tools in this endeavor are HMMER, BLAST, and specialized domain databases, which provide complementary approaches for remote homology detection and functional annotation.

This guide provides an objective performance comparison of these fundamental tools, with a specific focus on their application in profiling the diverse landscape of NBS domain genes across plant species. Understanding the relative strengths and limitations of these methods is essential for researchers investigating plant immunity mechanisms, developing disease-resistant crops, and exploring the evolutionary dynamics of plant immune systems. We present experimental data and detailed methodologies to inform tool selection for specific research scenarios in comparative plant genomics.

BLAST (Basic Local Alignment Search Tool)

BLAST operates on the principle of local sequence alignment, identifying regions of local similarity between sequences without requiring global alignment. Its heuristic approach makes it fast and practical for searching large databases. PSI-BLAST (Position-Specific Iterated BLAST) extends this capability by building a position-specific scoring matrix from significant hits in an initial search and iteratively searching the database with this profile, enhancing sensitivity to distant relationships.

HMMER (Profile Hidden Markov Models)

HMMER employs probabilistic profile hidden Markov models to represent sequence families and identify remote homologs. Unlike BLAST's pairwise approach, HMMER builds statistical models of multiple sequence alignments, capturing conserved patterns, insertions, and deletions across entire protein domains. This makes it particularly powerful for identifying divergent members of protein families based on subtle conserved motifs.

Domain Databases (Pfam, InterPro, CDD)

Domain databases provide curated multiple sequence alignments, HMMs, and functional annotations for protein domains and families. The Pfam database, for instance, uses HMMER software for its domain annotations and is particularly valuable for identifying NBS domains and other structural motifs in protein sequences through domain architecture analysis.

Table 1: Core Bioinformatics Tools for NBS Domain Gene Analysis

Tool	Primary Methodology	Key Strength	Typical Use Case in NBS Research
BLAST	Local sequence alignment via heuristic search	Speed, familiarity, widespread use	Initial identification of obvious NBS homologs; quick database searches
PSI-BLAST	Position-specific scoring matrix with iteration	Improved detection of distant relationships	Finding divergent NBS genes when initial BLAST fails
HMMER	Profile hidden Markov models	Sensitivity to very distant homologs; domain detection	Comprehensive identification of NBS domain genes; building custom gene families
Pfam/Domain DBs	Curated HMMs and alignments	Expert-curated models; standardized annotations	NBS domain identification and classification; functional inference

Performance Comparison: Experimental Data and Benchmarks

Remote Homology Detection

A systematic comparison published in Nucleic Acids Research evaluated the performance of HMMER and SAM (another profile HMM package) against PSI-BLAST and other non-HMM methods. The study found that profile HMM methods generally outperformed pairwise methods in detecting remote homology, with the quality of multiple sequence alignments used to build models being the most critical factor affecting overall performance [41].

In tests against the nrdb90 non-redundant database using globin and cupredoxin families, profile HMM methods demonstrated superior detection capabilities for distantly related sequences. The SAM package with its T99 iterative database search procedure performed better than the most recent version of PSI-BLAST at the time of the study. However, the scoring of PSI-BLAST profiles was reported to be more than 30 times faster than scoring of SAM models [41].

Computational Efficiency

The computational requirements of these tools vary significantly, impacting their practicality for large-scale genomic analyses. In the same comparative study, HMMER was found to be between one and three times faster than SAM when searching databases larger than 2000 sequences, with SAM being faster on smaller databases [41]. For typical NBS domain analyses involving thousands of sequences across multiple plant genomes, these efficiency considerations become important factors in tool selection.

Table 2: Performance Metrics for Bioinformatics Tools in Family-Wide Analysis

Performance Metric	BLAST	PSI-BLAST	HMMER	Domain Databases
Remote Homology Sensitivity	Moderate	Good	Excellent	Varies by curation
Speed	Fast	Moderate (faster scoring)	Slower model building, faster than SAM	Fast searching
Multiple Sequence Alignment Dependency	Not applicable	Moderate dependency	High dependency (critical factor)	Pre-curated models
E-value Accuracy	Good	Good	Comparable to HMMER	Dependent on underlying method
Low Complexity Masking	Effective	Effective	Effective using null models	Not applicable

Workflow Integration for NBS Domain Gene Analysis

Recommended Integrated Approach

A robust workflow for comparative analysis of NBS domain genes across plant species leverages the complementary strengths of these tools:

• Initial Screening with BLAST: Use BLAST against reference databases to identify clear homologs of known NBS domain genes as seeds for further analysis.

• Domain Identification with HMMER/Pfam: Search protein sequences against Pfam NBS models (e.g., NB-ARC domain, PF00931) using HMMER to confirm domain architecture and identify divergent family members.

• Custom Model Building with HMMER: For specialized analyses, build custom HMMs from high-quality multiple sequence alignments of identified NBS genes.

• Iterative Search with PSI-BLAST: Use PSI-BLAST to identify additional divergent family members that may have been missed in initial searches.

• Classification and Architecture Analysis: Use domain database annotations to classify NBS genes into subfamilies (TNL, CNL, etc.) based on domain architecture and identify species-specific structural patterns.

Experimental Protocol for NBS Gene Identification

The following detailed methodology has been successfully applied in large-scale comparative analyses of NBS domain genes:

Step 1: Sequence Data Collection

• Obtain proteome files for target plant species from public databases (Phytozome, NCBI, Plaza)
• For the NBS gene study across 34 species covering mosses to monocots and dicots, researchers used latest genome assemblies from publicly available databases [4]

Step 2: NBS Domain Identification

• Use HMMER-based search with Pfam NBS models (NB-ARC domain, PF00931)
• Apply PfamScan.pl HMM search script with default e-value (1.1e-50) using background Pfam-A_hmm model [4]
• Consider all genes having NB-ARC domain as NBS genes for further analysis

Step 3: Domain Architecture Classification

• Identify additional associated decoy domains through domain architecture analysis
• Classify genes into architectural classes (NBS, NBS-LRR, TIR-NBS, TIR-NBS-LRR, etc.) following established classification systems [4]
• Document both classical and species-specific structural patterns

Step 4: Orthogroup Analysis

• Use OrthoFinder v2.5.1 package with DIAMOND tool for fast sequence similarity searches
• Perform clustering using MCL clustering algorithm
• Identify core orthogroups and species-specific expansions

Step 5: Evolutionary Analysis

• Perform multiple sequence alignment using MAFFT 7.0
• Construct phylogenetic trees using maximum likelihood algorithm in FastTreeMP with 1000 bootstrap replicates [4]

NBS Domain Gene Analysis Workflow

Case Study: Large-Scale NBS Domain Analysis Across Plant Species

Experimental Framework and Results

A comprehensive study analyzing NBS domain genes across 34 plant species provides a practical example of this integrated approach [4]. Researchers identified 12,820 NBS-domain-containing genes, classifying them into 168 classes with several novel domain architecture patterns. The analysis revealed significant diversity among plant species, with both classical (NBS, NBS-LRR, TIR-NBS, TIR-NBS-LRR) and species-specific structural patterns (TIR-NBS-TIR-Cupin1-Cupin1, TIR-NBS-Prenyltransf, Sugar_tr-NBS).

The orthogroup analysis revealed 603 orthogroups, with some core (most common orthogroups) and unique (highly species-specific) orthogroups showing evidence of tandem duplications. Expression profiling demonstrated putative upregulation of specific orthogroups (OG2, OG6, OG15) in different tissues under various biotic and abiotic stresses in susceptible and tolerant plants to cotton leaf curl disease (CLCuD) [4].

Functional Validation

The study extended beyond bioinformatics prediction to functional validation through virus-induced gene silencing (VIGS) of a candidate NBS gene (GaNBS from OG2) in resistant cotton, demonstrating its putative role in virus tittering [4]. This validation highlights the importance of connecting computational predictions with experimental verification in planta.

Table 3: Key Research Reagent Solutions for NBS Domain Gene Studies

Reagent/Resource	Function/Purpose	Example Sources/Platforms
Genome Assemblies	Reference sequences for gene prediction and annotation	NCBI, Phytozome, Plaza Genome Databases
Pfam HMM Models	Curated profile HMMs for domain identification	Pfam database (NB-ARC: PF00931)
OrthoFinder	Orthogroup inference and comparative genomics	Software package for orthology assignment
MAFFT	Multiple sequence alignment for phylogenetic analysis	Alignment software package
FastTreeMP	Phylogenetic tree construction	Maximum likelihood tree building algorithm
RNA-seq Data	Expression profiling across tissues and conditions	IPF Database, CottonFGD, Cottongen
VIGS Vectors	Functional validation through gene silencing	TRV-based vectors for plant functional genomics

Emerging Approaches and Future Directions

While HMMER, BLAST, and domain databases remain foundational for NBS domain gene analysis, emerging approaches are expanding the bioinformatics toolkit. Deep learning-based functional representation methods like FRoGS (Functional Representation of Gene Signatures) show promise in enhancing target prediction by capturing functional relationships beyond simple sequence identity [42]. Similarly, AlphaFold 3 enables prediction of protein complex structures, potentially illuminating interactions between NBS domain proteins and their signaling partners [43].

The field continues to advance with improvements in genomic resources. As noted in a recent review of medicinal plant genomics, while over 400 genomes from 203 medicinal plants have been sequenced, challenges remain in assembly and annotation quality, with only 11 gapless telomere-to-telomere assemblies available as of February 2025 [44]. Enhanced genomic resources will further improve the accuracy of NBS domain gene annotation across diverse plant taxa.

NBS Domain Protein Architecture and Function

The integrated use of HMMER, BLAST, and domain databases provides a powerful framework for comparative analysis of NBS domain genes across plant species. Performance data demonstrates that while HMMER offers superior sensitivity for detecting remote homologs, BLAST provides complementary strengths in speed and practicality. The selection of appropriate tools should be guided by specific research objectives, with profile HMM methods being particularly valuable for comprehensive identification of divergent NBS domain genes, and BLAST-based approaches offering efficient solutions for initial screening and rapid database searches.

For researchers investigating the evolution of plant immune systems or developing disease-resistant crops, this integrated bioinformatics workflow enables robust identification, classification, and functional prediction of NBS domain genes across diverse plant taxa. As genomic resources continue to expand and new computational approaches emerge, these foundational tools will remain essential components of the plant genomics toolkit.

In the field of comparative genomics of NBS domain genes across plant species, accurately identifying and classifying nucleotide-binding site (NBS) domains is fundamental to understanding plant disease resistance mechanisms. The NBS domain is a conserved region found in numerous plant disease resistance (R) genes, particularly in the prominent NBS-LRR (Nucleotide-Binding Site Leucine-Rich Repeat) class of proteins that play critical roles in innate immunity [19] [13]. Researchers primarily rely on computational tools to identify these domains within protein sequences, with InterProScan, Pfam, and the Conserved Domains Database (CDD) representing three of the most widely used resources. These tools help annotate protein sequences by identifying domains and functional sites, but they differ in their underlying methods, coverage, and performance. This guide provides an objective comparison of these tools specifically for NBS domain validation, supported by experimental data and detailed protocols relevant to plant genomics research.

InterProScan functions as a meta-resource that integrates multiple protein signature databases, including both Pfam and CDD, into a unified framework [45]. It consolidates and cross-references annotations to produce a comprehensive overview of protein families, domains, and functional sites, reducing redundancy and enhancing annotation robustness [45]. Each integrated signature is assigned a unique InterPro entry; for example, signatures from CDD, PROSITE, Pfam, and SMART representing the same biological entity are consolidated into a single InterPro entry [45].

Pfam is a specialized database of protein families and domains, each represented by multiple sequence alignments and hidden Markov models (HMMs) [45]. Recently, the Pfam website has been decommissioned and its data fully integrated into the InterPro resource, making InterPro the primary access point for Pfam data [45].

CDD (Conserved Domains Database) provides protein domain annotations based on multiple sequence alignments of conserved domains, with a strong emphasis on 3D structure information [45]. It is one of the 13 member databases currently integrated into InterPro [45].

Table 1: Fundamental Characteristics of the Protein Classification Tools

Tool	Primary Classification Method	Integration Status in InterPro	Update Frequency
InterProScan	Integrated meta-scanner (13 databases)	N/A (Parent resource)	8-week release cycle [45]
Pfam	Hidden Markov Models (HMMs) [45]	Fully integrated (96.3% of signatures) [45]	Version 37.0 (as of 2024) [45]
CDD	Position-Specific Scoring Matrices [45]	Partially integrated (26.0% of signatures) [45]	Version 3.20 (as of 2024) [45]

Performance Comparison and Coverage Analysis

The performance of these tools varies significantly in terms of sequence coverage and domain integration. As of late 2024, InterPro provides annotations for over 200 million sequences, covering 81.8% of UniProtKB and 81.0% of UniParc sequences [45]. At the residue level, InterPro entries cover approximately 74% of all amino acids in UniProtKB, with member databases pending integration covering an additional 4.2% [45].

However, the integration rates of member databases into InterPro vary considerably. As shown in Table 1, Pfam exhibits excellent integration with 96.3% of its signatures incorporated into InterPro entries, while CDD shows much lower integration at only 26.0% [45]. This disparity suggests that using CDD through InterProScan may provide incomplete coverage compared to accessing CDD directly, particularly for specialized domains like NBS.

Limitations in Domain Detection

A critical study evaluating the capability of protein databases to identify specific functional domains revealed significant limitations. When analyzing 78 putative bacterial lipase sequences, InterProScan predicted lipase family membership for only 18 sequences (23%) and failed to predict any protein family membership for 41 sequences (53%) [46]. Furthermore, the study noted that different scanning tools produced inconsistent and non-consensus predictions for the same sequences, highlighting that even an integrated tool like InterProScan may miss genuine domain features present in specialized databases [46].

These findings are particularly relevant for NBS domain researchers, as they demonstrate that reliance on a single tool, even a comprehensive one like InterProScan, may yield incomplete annotations, especially for novel or taxonomically restricted domains.

Table 2: Performance Metrics for Protein Domain Annotation Tools

Performance Metric	InterProScan	Pfam (via InterPro)	CDD (via InterPro)
Member Database Integration	100% (by definition)	96.3% [45]	26.0% [45]
UniProtKB Sequence Coverage	81.8% (201 million+ sequences) [45]	Part of InterPro coverage	Part of InterPro coverage
Case Study Detection Rate	23% (lipase features) [46]	Information not available in search results	Information not available in search results
Key Strength	Comprehensive, non-redundant annotations	High-quality HMMs for families	Structural domain perspective

Experimental Protocols for NBS Domain Validation

Genome-Wide Identification of R-Genes

The following protocol, adapted from cowpea and potato genomic studies, outlines a standard workflow for identifying and validating NBS domains in plant genomes [19] [13]:

• Sequence Acquisition and Preparation: Obtain protein sequences of interest from whole-genome sequencing assemblies or transcriptome data. For cowpea R-gene identification, researchers used a hybrid assembly approach combining Illumina and Nanopore sequencing technologies to generate a high-quality genome assembly [19].

• Initial Domain Scanning:

  • Process all protein sequences through InterProScan to identify candidate R-genes containing NBS domains.
  • Use default parameters for domain detection, which will leverage integrated member databases including Pfam and CDD.
  • Extract sequences with NBS domain hits for further validation.

• Secondary Validation with Individual Tools:

  • Process the candidate sequences through CDD's standalone tools (if available) to identify any additional NBS domains not detected through InterProScan.
  • Similarly, process sequences using Pfam's HMM models directly, though note that Pfam is now primarily accessed through InterPro.

• Manual Curation and Classification:

  • Classify validated NBS-containing genes into subclasses (e.g., CNL, TNL) based on associated domains like coiled-coil (CC) or Toll/interleukin-1 receptor (TIR) domains [19].
  • Verify domain architecture through multiple tools to resolve conflicting annotations.

The following workflow diagram illustrates the sequential steps for this experimental protocol:

NBS-Tag Profiling for Population Studies

For comparative genomics across multiple cultivars or plant species, NBS-tag profiling provides a targeted approach [13]:

• Primer Design: Design degenerate PCR primers targeting conserved motifs within the NBS domain (P-loop, Kinase-2, and GLPL motifs) [13].

• Library Preparation and Sequencing: Amplify NBS tags from genomic DNA using these primers and sequence using high-throughput platforms (e.g., Illumina HiSeq) [13].

• Read Mapping and Variant Calling: Map sequenced NBS tags to a reference genome and identify single nucleotide polymorphisms (SNPs) and insertions/deletions (indels) within NBS domains.

• Functional Annotation: Annotate polymorphic NBS domains using InterProScan, CDD, and Pfam to assess potential functional impacts of identified variations.

Table 3: Key Research Reagents and Computational Tools for NBS Domain Analysis

Resource	Type	Primary Function in NBS Research	Access Information
InterProScan	Software Tool	Integrated protein domain and family annotation [45]	https://www.ebi.ac.uk/interpro [45]
Pfam Database	Protein Family Database	Curated HMMs for identifying NBS domains and other protein families [45]	Accessed via InterPro [45]
CDD Database	Domain Database	Provides conserved domain annotations with structural information [45]	https://www.ncbi.nlm.nih.gov/cdd/ [45]
UniProtKB	Protein Sequence Database	Standard repository of reviewed and unreviewed protein sequences [45]	https://www.uniprot.org/ [45]
PRGminer	Specialized Tool	Deep learning-based prediction of plant resistance genes [47]	https://kaabil.net/prgminer/ [47]
Degenerate PCR Primers	Wet Lab Reagent	Amplification of NBS domain fragments from genomic DNA [13]	Custom-designed for conserved NBS motifs [13]

For researchers validating NBS domains in plant species, the combined use of InterProScan, CDD, and Pfam provides complementary advantages. InterProScan offers the most efficient and comprehensive initial scan, leveraging its integrated database structure. However, given CDD's low integration rate (26.0%) and the documented limitations of protein classifiers in detecting all genuine domain features, supplementing InterProScan with direct CDD analysis is strongly recommended for critical NBS domain validation work. This multi-tool approach is particularly crucial for identifying novel NBS domains in non-model plant species or those with limited prior characterization, ensuring maximal detection sensitivity and annotation accuracy in comparative genomics studies of plant disease resistance genes.

Nucleotide-binding site (NBS) genes constitute one of the largest and most critical disease resistance (R) gene families in plants, playing indispensable roles in innate immune responses against diverse pathogens [48] [33]. These genes typically encode proteins characterized by a conserved NBS domain alongside C-terminal leucine-rich repeats (LRRs) and variable N-terminal domains that define their subfamily classification: coiled-coil (CC-NBS-LRR or CNL), Toll/Interleukin-1 receptor (TIR-NBS-LRR or TNL), or Resistance to Powdery Mildew8 (RPW8-NBS-LRR or RNL) [48] [4]. The NBS gene family exhibits remarkable diversity across plant genomes, with copy numbers ranging from fewer than 100 to over 1,000 members, reflecting dynamic evolutionary processes shaped by host-pathogen co-evolution [4] [33].

Orthogroup analysis has emerged as a fundamental methodology in comparative genomics, enabling researchers to identify groups of genes descended from a single ancestral gene in a common ancestor of the species being compared [49] [50]. This approach provides an evolutionarily coherent framework for investigating gene family evolution across multiple species, overcoming limitations of pairwise orthology inference methods that struggle with complex genomic histories involving duplications and losses [49] [51]. For NBS genes, which are frequently organized in tandem arrays and subject to frequent duplication events, orthogroup analysis offers particular value for tracing evolutionary patterns, identifying conserved gene clusters, and understanding the genomic basis of disease resistance mechanisms [48] [52].

This guide provides a comprehensive comparison of experimental approaches, computational tools, and analytical frameworks for conducting orthogroup analysis of NBS genes across plant species, with emphasis on practical implementation and interpretation of results within the context of comparative genomics research.

Methodological Framework for NBS Gene Identification and Classification

Domain Architecture and Gene Identification Protocols

The initial critical step in orthogroup analysis involves the comprehensive identification of NBS-encoding genes across target genomes. This process typically employs a dual search strategy combining homology-based and profile-based methods to ensure maximum coverage [9] [4]. The standard protocol utilizes Hidden Markov Model (HMM) searches with the conserved NB-ARC domain (Pfam accession: PF00931) as query, complemented by BLAST or BLASTp analyses against reference NBS protein sequences from well-annotated genomes such as Arabidopsis thaliana, Oryza sativa, or other relevant species [48] [9].

For HMM searches, the recommended parameters include using the PfamScan.pl script with default e-value (1.1e-50) against the Pfam-A_hmm model, retaining all sequences containing the NB-ARC domain for subsequent analysis [4]. For BLAST searches, stringent E-value cutoffs of 1e-10 or lower should be applied to minimize false positives [9]. Candidate sequences identified through these methods must undergo validation through domain architecture analysis using tools such as InterProScan or NCBI's Batch CD-Search to confirm the presence of characteristic NBS domain structures and additional domains (CC, TIR, RPW8, LRR) that facilitate functional classification [48] [9].

Table 1: Standard Protocols for NBS Gene Identification

Method Type	Key Tools	Parameters	Validation Approach
HMM Search	HMMER/PfamScan	Pfam PF00931, E-value 1.1e-50	Domain confirmation with InterProScan
BLAST Search	BLAST+/DIAMOND	E-value ≤1e-10, reference sequences	Reciprocal best hits
Domain Analysis	InterProScan, NCBI CD-Search	E-value ≤1e-5	Architecture classification

Classification and Structural Characterization

Following identification, NBS genes are classified into subfamilies based on their N-terminal domains and overall domain architecture [48] [4]. This classification employs a combination of automated domain annotation and motif analysis. The MEME suite can be utilized for predicting conserved motifs within NBS domains with the motif number typically set to 10 while maintaining default parameters [9]. Gene structures are subsequently analyzed through GSDS 2.0 (Gene Structure Display Server), providing visual representation of exon-intron organization that may reveal evolutionary relationships [9].

Additional characterization includes promoter analysis using PlantCARE to identify cis-acting regulatory elements in the 2000 bp upstream regions, revealing potential regulatory patterns associated with defense responses [9]. Subcellular localization predictions can be performed using WoLF PSORT, providing insights into potential functional specialization [9]. This comprehensive characterization facilitates not only functional predictions but also informs the orthogroup analysis by highlighting structural conservation beyond sequence similarity.

Comparative Analysis of Orthology Inference Algorithms

Algorithm Performance Benchmarking

Selecting appropriate orthology inference algorithms is crucial for robust orthogroup analysis. Multiple tools have been developed with different underlying methodologies, each with distinct strengths and limitations for analyzing complex gene families like NBS genes [49] [51]. A recent comparative study evaluating four orthology inference algorithms—OrthoFinder, SonicParanoid, Broccoli, and OrthNet—on Brassicaceae genomes revealed that while all methods showed general consistency, significant differences emerged in handling complex genomic histories [49].

OrthoFinder consistently demonstrates high accuracy in ortholog inference, outperforming other methods on standard benchmarks. In comprehensive tests using the Quest for Orthologs benchmark dataset, OrthoFinder was 3-24% (SwissTree) and 2-30% (TreeFam-A) more accurate than competing methods [50]. This performance advantage stems from its phylogenetic approach to orthology inference, which distinguishes variable sequence evolution rates from true phylogenetic relationships, thereby reducing both false-positive and false-negative errors [50]. The algorithm employs a multi-step process involving orthogroup inference, gene tree construction, rooted species tree inference, and duplication-loss-coalescence analysis to delineate orthologs and paralogs [50].

SonicParanoid and Broccoli also demonstrate strong performance, with SonicParanoid employing a graph-based inference algorithm modified from the InParanoid approach, while Broccoli uses tree-based methods with network analyses to determine orthology relationships [49]. All three programs effectively account for gene length biases before clustering proteins based on sequence similarity. OrthNet, which incorporates synteny information through the CLfinder workflow, generally produced more divergent results but provided valuable information about gene colinearity [49].

Table 2: Orthology Inference Algorithm Comparison

Algorithm	Methodology	Strengths	Limitations	Best Use Cases
OrthoFinder	Phylogenetic tree-based	Highest accuracy, comprehensive outputs	Computationally intensive	Reference-quality analyses
SonicParanoid	Graph-based	Fast, efficient for large datasets	Limited phylogenetic context	High-throughput screening
Broccoli	Tree-based with network analysis	Balanced approach	Moderate computational demand	General comparative studies
OrthNet	Synteny-aware	Colinearity information	Divergent results	Ancestral genome reconstruction

Impact of Genomic Complexity on Orthology Inference

The performance of orthology inference algorithms is significantly influenced by genomic complexity, particularly whole-genome duplication events and varying ploidy levels [49]. Studies comparing orthogroup inference in diploid versus polyploid Brassicaceae species revealed that diploid sets exhibited a higher proportion of identical orthogroups, while sets including mesopolyploids and recent allohexaploids showed lower proportions of identically composed orthogroups, though average similarity degrees remained comparable [49].

This has important implications for NBS gene analysis, as these genes frequently reside in complex genomic regions with elevated duplication rates. Phylogeny-aware methods like OrthoFinder generally outperform synteny-based approaches for orthology detection in such dynamic genomic contexts [51]. However, synteny-based approaches (e.g., Roary, PanOCT) provide advantages for identifying vertically transmitted members of mobile gene families when applied to closely related species with conserved gene order [51].

Experimental Design and Workflow for NBS Orthogroup Analysis

Integrated Analysis Pipeline

A robust workflow for NBS orthogroup analysis integrates multiple computational steps from gene identification through evolutionary interpretation. The following diagram illustrates a comprehensive pipeline:

Implementation Considerations

Successful implementation of orthogroup analysis requires careful consideration of taxonomic sampling and data quality. Studies investigating NBS gene evolution across land plants have demonstrated that including species representing key evolutionary nodes (bryophytes, lycophytes, basal angiosperms, monocots, and eudicots) enables more accurate reconstruction of evolutionary trajectories [4] [33]. Genome quality assessment using BUSCO scores should precede analysis, with preference given to assemblies with >90% completeness for core gene sets [48] [9].

For orthogroup inference itself, OrthoFinder implementation typically begins with all-vs-all sequence similarity searches using DIAMOND or BLAST, followed by orthogroup inference using the Markov Clustering algorithm (MCL) [50]. The resulting orthogroups then undergo gene tree inference using fast phylogenetic methods such as DendroBLAST or more rigorous approaches like MAFFT alignment followed by FastTree or RAxML tree inference [4] [50]. The species tree is inferred from the complete set of gene trees using statistical approaches, which subsequently enables accurate rooting of gene trees and identification of duplication events [50].

Case Studies in Plant Lineages

Sapindaceae Family Analysis

A comprehensive analysis of NBS-encoding genes in three Sapindaceae species (Xanthoceras sorbifolium, Dinnocarpus longan, and Acer yangbiense) revealed distinct evolutionary patterns driven by species-specific duplication and loss events [48]. Researchers identified 180, 568, and 252 NBS-encoding genes in these species respectively, with uneven chromosomal distribution and predominant organization in tandem arrays rather than as singletons [48].

Phylogenetic reconstruction classified these genes into three monophyletic clades (RNL, TNL, and CNL) distinguished by amino acid motifs [48]. Analysis of ancestral genes revealed that the NBS-encoding genes in these three genomes derived from 181 ancestral genes (3 RNL, 23 TNL, and 155 CNL), with dynamic evolutionary patterns emerging post-speciation [48]. X. sorbifolium exhibited an evolutionary pattern of "first expansion and then contraction," while A. yangbiense and D. longan showed a "first expansion followed by contraction and further expansion" pattern, with D. longan experiencing particularly strong recent expansion potentially corresponding to adaptation to diverse pathogens [48].

Asparagus Genus Investigation

A comparative analysis of NLR genes across garden asparagus (Asparagus officinalis) and its wild relatives (A. kiusianus and A. setaceus) demonstrated how domestication has influenced NBS gene repertoire [9]. The study identified 63, 47, and 27 NLR genes in A. setaceus, A. kiusianus, and A. officinalis respectively, revealing marked contraction associated with domestication [9]. Orthologous gene analysis identified 16 conserved NLR gene pairs between A. setaceus and A. officinalis, representing NLR genes preserved during domestication [9].

Functional investigations coupled with this orthogroup analysis revealed that despite pathogen challenge, most preserved NLR genes in cultivated asparagus showed unchanged or downregulated expression, suggesting potential functional impairment in disease resistance mechanisms as a consequence of selection for yield and quality traits [9]. This case study exemplifies how orthogroup analysis can reveal both quantitative and qualitative changes in NBS gene complement associated with evolutionary processes.

Coffee Tree Resistance Locus Evolution

Investigation of the SH3 locus conferring resistance to coffee leaf rust in Coffea arabica provided insights into the evolution of a specific NBS gene cluster [52]. Sequence analysis of the SH3 region in three coffee genomes (Ea and Ca subgenomes from allotetraploid C. arabica and Cc genome from diploid C. canephora) revealed 5, 3, and 4 R genes respectively, all belonging to a CC-NBS-LRR (CNL) family exclusively found at the SH3 locus [52].

Orthology relationship determination enabled researchers to trace duplication/deletion events shaping the SH3 locus, revealing that the origin of most SH3-CNL copies predated speciation within Coffea [52]. The SH3-CNL family evolution followed the birth-and-death model, with gene conversion between paralogs, inter-subgenome sequence exchanges, and positive selection acting as major evolutionary forces [52]. This case highlights how orthogroup analysis at the micro-evolutionary scale can elucidate mechanisms driving resistance gene evolution.

Computational Tools and Databases

Table 3: Essential Resources for NBS Orthogroup Analysis

Resource Category	Specific Tools/Databases	Primary Function	Application Notes
Genome Databases	Phytozome, PLAZA, NCBI Genome	Access to genomic sequences and annotations	PLAZA integrates comparative genomics data for 25+ plant species
Orthology Inference	OrthoFinder, SonicParanoid, Broccoli	Orthogroup identification	OrthoFinder provides highest accuracy in benchmark tests
Domain Analysis	InterProScan, Pfam, CDD	Protein domain identification	Critical for NBS gene classification and validation
Sequence Alignment	MAFFT, Clustal Omega	Multiple sequence alignment	MAFFT generally preferred for large datasets
Phylogenetic Analysis	FastTree, RAxML, MEGA	Gene tree construction	FastTree balances speed and accuracy for large orthogroups
Visualization	TBtools, iTOL, GSDS	Results visualization and interpretation	TBtools specifically designed for genomic data

Experimental Validation Approaches

Orthogroup analysis generates hypotheses about gene function and evolution that frequently require experimental validation. Several key approaches enable such validation:

Expression profiling under pathogen challenge or stress conditions provides insights into functional conservation. Studies in cotton have demonstrated differential expression of specific orthogroups (OG2, OG6, OG15) in response to cotton leaf curl disease between susceptible and tolerant accessions [4]. RNA-seq data analysis across tissues and stress conditions can reveal expression conservation among orthologs, supporting functional predictions based on orthogroup membership [4].

Functional characterization through virus-induced gene silencing (VIGS) has proven valuable for validating NBS gene function. Silencing of GaNBS (OG2) in resistant cotton demonstrated its putative role in virus tittering, confirming predictions from orthogroup analysis [4]. Similarly, protein-ligand and protein-protein interaction studies can reveal conserved interaction patterns, such as strong interaction between putative NBS proteins and ADP/ATP or core proteins of the cotton leaf curl disease virus [4].

Genetic variation analysis between resistant and susceptible genotypes can identify functionally significant polymorphisms within NBS orthogroups. Studies comparing tolerant (Mac7) and susceptible (Coker 312) Gossypium hirsutum accessions identified numerous unique variants in NBS genes (6583 in Mac7 versus 5173 in Coker312), highlighting potential functional differences [4].

Interpretation and Evolutionary Analysis of Results

Evolutionary Patterns and Inferences

Orthogroup analysis of NBS genes across multiple plant lineages has revealed distinctive evolutionary patterns that reflect different adaptive strategies. Studies across diverse angiosperms have shown that CNL genes generally exhibit gradual expansion patterns, with intense expansion corresponding to fungal diversity explosions, while RNL genes typically maintain low copy numbers due to conserved functions [48] [33]. The evolutionary history of NBS genes is characterized by frequent birth-and-death evolution, with lineage-specific expansions and contractions driven by pathogen pressure [48] [52].

Analysis of 12,820 NBS-domain-containing genes across 34 species from mosses to monocots and dicots identified 168 classes with both classical and species-specific domain architecture patterns [4]. Orthogroup analysis revealed 603 orthogroups with core (widely distributed) and unique (species-specific) orthogroups showing evidence of tandem duplications [4]. These patterns reflect the dynamic nature of NBS gene evolution, with different plant lineages employing distinct strategies for maintaining disease resistance gene repertoires.

Technical Considerations and Limitations

While orthogroup analysis provides powerful insights into NBS gene evolution, several technical considerations merit attention. The choice of clustering criterion significantly impacts downstream analyses, with phylogeny-aware methods (OrthoFinder, panX) and synteny-based approaches (Roary) producing meaningfully different results for certain pangenome features [51]. This variability can exceed ecological and phylogenetic effect sizes for some pangenome features, necessitating careful method selection aligned with research objectives [51].

Gene annotation quality represents another critical factor, as fragmented or incomplete gene models can disrupt orthogroup inference. Integration of transcriptomic evidence to refine gene models before orthogroup analysis significantly improves results [9] [4]. Additionally, taxonomic sampling density influences evolutionary inferences, with sparse sampling potentially leading to inaccurate reconstruction of duplication events and orthology relationships [49] [50].

Finally, the complex genomic architecture of NBS genes—frequent tandem arrays, sequence similarity, and gene conversion—presents particular challenges for orthology inference algorithms. Integration of multiple approaches, including synteny information and phylogenetic analysis, provides the most robust results for these challenging but biologically crucial gene families [49] [52].

In plant genomes, nucleotide-binding site (NBS) domain genes encode a critical class of immune receptors that confer resistance to diverse pathogens. These genes exhibit remarkable structural diversity and species-specific expansion patterns across land plants, with over 12,800 NBS-domain-containing genes identified from mosses to monocots and dicots [4]. While coding sequence variation contributes to pathogen recognition specificity, the regulation of these defense genes is equally crucial for mounting effective immune responses. Promoter and cis-regulatory element analysis provides a powerful framework for understanding how plants control the expression of their defense arsenal, connecting specific DNA sequence motifs to transcriptional outputs that determine resistance outcomes. This review integrates comparative genomic findings with experimental data to elucidate how regulatory sequences shape plant immunity through the coordinated expression of NBS domain genes, offering insights for engineering durable disease resistance in crop species.

Cis-Element Diversity in NBS Gene Promoters

Genome-Wide Identification of Defense-Related Cis-Elements

Comprehensive analyses of promoter regions upstream of NBS domain genes have revealed an enrichment of cis-elements responsive to defense signals and phytohormones. In asparagus species, promoters of NLR genes contained "numerous cis-elements responsive to defense signals and phytohormones" [9]. Similar findings were reported in Nicotiana species, where analysis of 1500 bp promoter sequences upstream of NBS-LRR genes identified 29 shared types of regulatory elements, including four kinds unique to irregular-type NBS-LRR genes [53]. This conservation of regulatory architecture across species suggests fundamental principles in the transcriptional control of plant immunity.

The functional significance of these cis-elements was demonstrated in Lolium multiflorum, where the LmMYB1 gene promoter showed significantly increased expression under drought and ABA stress conditions [54]. This expression pattern correlated with the presence of ABA-responsive elements in the promoter region, highlighting how specific cis-elements directly mediate transcriptional responses to environmental stresses. Similarly, in cotton, expression profiling revealed putative upregulation of specific orthogroups (OG2, OG6, and OG15) in different tissues under various biotic and abiotic stresses in plants with varying susceptibility to cotton leaf curl disease [4].

Table 1: Experimentally Validated Cis-Elements in NBS Gene Promoters

Cis-Element	Consensus Sequence	Transcription Factor	Function in Defense	Experimental Validation
M1 (Caenorhabditis)	GAGACCY	Unknown	Germline development, oogenesis	Reporter constructs in transgenic C. elegans [55]
M2 (Caenorhabditis)	GYGCCTTT	Unknown	Germline development, oogenesis	Reporter constructs in transgenic C. elegans [55]
ABA-responsive element	Not specified	MYB transcription factors	Drought stress response	Expression analysis in Lolium multiflorum [54]
Defense-responsive elements	Not specified	Not specified	Pathogen response	Promoter analysis in asparagus NLR genes [9]

Structural Constraints in Cis-Regulatory Modules

Beyond simple presence/absence of cis-elements, their spatial organization exhibits remarkable constraints that reflect functional requirements. In Caenorhabditis elegans, a novel pair of cis-regulatory motifs (GAGACCY and GYGCCTTT) displays "extraordinary genomic traits" including highly specific order and orientation, with almost invariant spacing of either 16 or 19 bases between them [55]. This nearly combinatorial configuration, conserved across the Caenorhabditis genus but absent in other nematodes, represents an exceptional example of structural constraint in regulatory sequences.

The functional implications of such constrained architectures likely relate to the stereospecific requirements for transcription factor assembly on DNA. The fixed distances of 16 and 19 bases between the Caenorhabditis motifs correspond approximately to 1.5 and 1.8 turns of the DNA double helix, potentially positioning transcription factors on the same face of the DNA to facilitate protein-protein interactions [55]. Similar structural constraints may govern the organization of cis-elements regulating NBS gene expression in plants, though these spatial relationships remain less characterized.

Methodological Framework for Cis-Element Analysis

Computational Identification Pipelines

Standardized bioinformatic workflows have emerged for the systematic identification and characterization of cis-regulatory elements in plant genomes. The typical analytical pipeline begins with the extraction of promoter sequences, generally defined as 1500-2000 bp upstream of the start codon [9] [53]. These sequences are then subjected to cis-element analysis using specialized databases such as PlantCARE, which provides comprehensive annotation of known plant regulatory elements [9] [53].

For NBS gene families, identification typically employs a dual approach combining Hidden Markov Model (HMM) searches using the conserved NB-ARC domain (Pfam: PF00931) as query, followed by validation through domain architecture analysis using tools like InterProScan and NCBI's Batch CD-Search [9] [28]. This integrated methodology ensures comprehensive identification of NBS genes while minimizing false positives. The application of this pipeline in Nicotiana species successfully identified 1226 NBS genes across three genomes, revealing that 76.62% of members in Nicotiana tabacum could be traced back to parental genomes [28].

Table 2: Key Bioinformatics Tools for Promoter and Cis-Element Analysis

Tool Category	Specific Tools	Function	Key Parameters
Promoter Sequence Extraction	TBtools, BEDTools	Extract upstream sequences	Typically 1500-2000 bp upstream of ATG
Cis-Element Annotation	PlantCARE	Identify known regulatory elements	Database of plant cis-acting elements
Domain Identification	HMMER, InterProScan, CDD	Identify protein domains	HMM model PF00931 for NBS domain
Motif Discovery	MEME Suite	Discover novel motifs	E-value < 1e-5, motif count 10
Phylogenetic Analysis	MEGA, Clustal Omega	Evolutionary relationships	Maximum likelihood, 1000 bootstraps

Experimental Validation Approaches

Computational predictions require experimental validation to confirm regulatory function. Reporter constructs in transgenic systems represent the gold standard for functional validation of cis-elements. In C. elegans, promoter GFP reporters demonstrated that the identified motif pair functioned as bona fide cis-regulatory elements controlling germline development [55]. Similarly, in plants, virus-induced gene silencing (VIGS) has proven valuable for functional characterization, as demonstrated by the silencing of GaNBS (OG2) in resistant cotton, which validated its putative role in virus resistance [4].

Expression analyses under stress conditions provide additional functional insights. In Lolium multiflorum, quantitative expression profiling following drought stress and ABA treatment revealed significant induction of LmMYB1, implicating ABA-responsive elements in its promoter [54]. Similar approaches in asparagus showed that most preserved NLR genes in susceptible A. officinalis demonstrated either unchanged or downregulated expression following fungal challenge, indicating potential functional impairment in disease resistance mechanisms [9].

Signaling Pathways in Defense Gene Regulation

The regulation of NBS domain genes involves complex signaling networks that integrate pathogen perception with transcriptional reprogramming. The diagram below illustrates the primary signaling pathway connecting pathogen recognition to defense gene activation through cis-element interactions.

This integrated signaling network illustrates how both biotic and abiotic stress pathways converge to regulate NBS gene expression through transcription factor binding to specific cis-elements. The ABA-dependent pathway exemplifies how abiotic stress signaling can influence disease resistance through both direct transcriptional regulation and physiological adaptations like reduced stomatal density [54].

Comparative Evolution of Regulatory Sequences

Evolutionary Dynamics of NBS Gene Regulation

The regulatory sequences controlling NBS gene expression exhibit distinct evolutionary patterns compared to coding sequences. In asparagus species, comparative genomic analysis revealed "a marked contraction of the NLR gene repertoire from the wild species to the domesticated A. officinalis," with gene counts of 63, 47, and 27 NLR genes identified in A. setaceus, A. kiusianus, and A. officinalis, respectively [9]. This contraction during domestication was accompanied by altered expression patterns, where "the majority of preserved NLR genes in A. officinalis demonstrated either unchanged or downregulated expression following fungal challenge" [9].

Orthologous gene analysis identified 16 conserved NLR gene pairs between A. setaceus and A. officinalis, representing NLR genes preserved during the domestication process [9]. The differential expression of these orthologs suggests that regulatory changes, potentially in promoter regions, contribute significantly to domestication-associated susceptibility. This pattern of regulatory evolution mirrors observations in other plant species, where human selection for yield and quality traits often inadvertently compromises defense gene expression.

Species-Specific Cis-Regulatory Innovation

While core regulatory modules are conserved across plant lineages, species-specific innovations continually emerge. The study of NBS domain genes across 34 plant species revealed "several classical (NBS, NBS-LRR, TIR-NBS, TIR-NBS-LRR, etc.) and species-specific structural patterns (TIR-NBS-TIR-Cupin1-Cupin1, TIR-NBS-Prenyltransf, Sugar_tr-NBS etc.)" [4]. This diversity in domain architecture likely correlates with promoter sequence variation, enabling species-specific regulation of defense responses.

The Caenorhabditis motif pair exemplifies how novel regulatory modules can emerge within specific evolutionary lineages. This motif pair is "conserved among, and unique to, the entire Caenorhabditis genus" [55], indicating its recent evolutionary origin and lineage-specific functional importance. Similar genus-specific cis-regulatory innovations likely exist in plant genomes, contributing to the diversification of defense gene regulation across taxa.

The Scientist's Toolkit: Essential Research Reagents

Table 3: Key Research Reagents for Promoter and Cis-Element Analysis

Reagent Category	Specific Examples	Function/Application	Key Features
Bioinformatics Databases	PlantCARE, Pfam, NCBI CDD	Cis-element annotation, domain identification	Curated collections of regulatory elements and protein domains
HMM Models	PF00931 (NB-ARC domain)	Identification of NBS domain genes	Specificity for nucleotide-binding domain
Expression Validation Systems	Virus-Induced Gene Silencing (VIGS)	Functional characterization of NBS genes	Transient silencing without stable transformation
Reporter Constructs	GFP/GUS reporter fusions	Validation of promoter activity	Visual assessment of spatial expression patterns
Genomic Resources	Genome assemblies of model and crop plants	Comparative analysis	Reference sequences for ortholog identification

Promoter and cis-element analysis provides fundamental insights into the regulatory logic governing plant defense responses. The integration of computational predictions with experimental validation has revealed conserved principles of defense gene regulation, while also highlighting species-specific innovations that contribute to immunological diversity. The continued development of genomic resources and analytical tools will further enhance our understanding of how regulatory sequences evolve and function in plant immunity. This knowledge provides a critical foundation for future efforts to engineer disease-resistant crops through targeted manipulation of defense gene regulatory circuits.

The nucleotide-binding site-leucine-rich repeat (NBS-LRR) gene family constitutes one of the most critical components of plant immune systems, encoding intracellular receptors that recognize pathogen effector molecules and initiate defense responses [56] [4]. These genes represent the largest class of plant resistance (R) genes, with approximately 60% of cloned disease resistance genes belonging to this family [28]. Proteins encoded by NBS-LRR genes typically contain a conserved nucleotide-binding site (NBS) domain and a C-terminal leucine-rich repeat (LRR) region, with variable N-terminal domains categorizing them into subfamilies such as TNL (TIR-NBS-LRR), CNL (CC-NBS-LRR), and RNL (RPW8-NBS-LRR) [9] [53]. The NBS domain primarily mediates signal transduction [28], while the LRR domain is responsible for specific pathogen recognition [28].

NBS-LRR genes exhibit remarkable diversity across plant species, with significant variation in gene counts—from as few as 25 NLRs in the bryophyte Physcomitrella patens to over 2,000 in bread wheat (Triticum aestivum) [4]. This extensive diversity, coupled with complex expression patterns influenced by multiple signaling pathways and environmental factors, presents substantial challenges for functional characterization. In this context, machine learning approaches offer powerful tools for deciphering the relationship between NBS gene sequences, their expression patterns, and their functions in stress responses.

Comparative Genomics of NBS Genes Across Plant Species

Genomic Distribution and Evolutionary Dynamics

Table 1: NBS-LRR Gene Distribution Across Plant Species

Plant Species	Total NBS Genes	TNL	CNL	NL	RNL	Other	Reference
Nicotiana tabacum (Tobacco)	603	9	150	64	74	306	[28]
Nicotiana benthamiana	156	5	25	23	4	99	[53]
Asparagus officinalis (Garden asparagus)	27	Not specified	Not specified	Not specified	Not specified	Not specified	[9]
Asparagus setaceus	63	Not specified	Not specified	Not specified	Not specified	Not specified	[9]
Vigna unguiculata (Cowpea)	2,188 (total R-genes)	Not specified	Not specified	Not specified	Not specified	Not specified	[19]

The expansion and contraction of NBS gene families across plant species reveal fascinating evolutionary patterns influenced by both whole-genome duplication (WGD) and small-scale duplication events [4]. Comparative genomic analysis in asparagus species revealed a notable contraction of NLR genes from wild species to the domesticated A. officinalis, with gene counts of 63, 47, and 27 NLR genes identified in A. setaceus, A. kiusianus, and A. officinalis, respectively [9]. This reduction in gene repertoire during domestication suggests potential trade-offs between disease resistance and agricultural traits selected by humans.

In tobacco (Nicotiana tabacum), an allotetraploid formed through hybridization of N. sylvestris and N. tomentosiformis, approximately 76.62% of NBS members could be traced back to their parental genomes, demonstrating the impact of polyploidization on NBS gene family expansion [28]. Whole-genome duplication contributed significantly to this expansion, with the total number of NBS genes in N. tabacum (603) approximately equaling the combined total of its progenitors (279 in N. tomentosiformis and 344 in N. sylvestris) [28].

Structural Diversity and Classification Frameworks

NBS-LRR genes display considerable structural diversity, leading to their classification into multiple categories based on domain architecture:

• Typical NBS-LRRs: Contain all three major domains (N-terminal, NBS, and LRR)
• Irregular NBS-LRRs: Lack one or more domains, potentially functioning as adaptors or regulators [53]
• Additional categories: Include CC-NBS (CN), CC-NBS-LRR (CNL), TIR-NBS (TN), TIR-NBS-LRR (TNL), RPW8-NBS (RN), RPW8-NBS-LRR (RNL), and NBS-only (N) types [28] [53]

A comprehensive study identifying 12,820 NBS-domain-containing genes across 34 plant species classified them into 168 distinct classes with several novel domain architecture patterns, revealing significant diversity across plant species [4]. This extensive structural variation underpins the functional diversification of NBS genes and provides a rich feature set for machine learning algorithms to exploit in function prediction.

Machine Learning Framework for NBS Gene Function Prediction

Feature Extraction and Dataset Construction

Table 2: Feature Categories for Machine Learning Models Predicting NBS Gene Function

Feature Category	Specific Features	Data Source	Prediction Relevance
Sequence-Based Features	Domain architecture, motif composition, conserved residues (P-loop, GLPL, MHD, Kinase 2), physicochemical properties	Genome sequencing, multiple sequence alignment	Structural-functional relationships, nucleotide binding specificity
Evolutionary Features	Orthogroup membership, synteny relationships, duplication history, selection pressure (Ka/Ks ratios)	Comparative genomics, phylogenetic analysis	Functional conservation, evolutionary constraints
Expression Features	Basal expression levels, induction kinetics under stress, tissue-specificity, alternative splicing	RNA-seq, microarray data	Stress responsiveness, spatiotemporal functionality
Epigenetic Features	DNA methylation patterns, histone modifications, chromatin accessibility	ChIP-seq, bisulfite sequencing	Regulatory mechanisms, expression potential
Promoter Features	Cis-regulatory elements (SA, JA, ABA responsiveness, stress-related elements)	Promoter analysis, footprinting	Regulatory logic, signaling pathway integration

The foundation of effective machine learning models for NBS gene function prediction lies in comprehensive feature extraction. The promoter regions of NBS genes contain numerous cis-elements responsive to defense signals and phytohormones [9], which can be identified using tools like PlantCARE [53]. For instance, analysis of the soybean SRC4 promoter identified 12 regulatory elements, including salicylic acid (SA)-responsive elements, which proved critical for understanding its transcriptional regulation [56].

Expression quantitative trait loci (eQTL) mapping combined with stress-responsive expression profiling provides valuable features for predicting gene function under specific environmental conditions. Studies have demonstrated that NBS genes show distinct expression patterns under various stress conditions, with some genes exhibiting broad-spectrum responsiveness [4] [57].

Algorithm Selection and Model Architectures

Multiple machine learning approaches can be employed for predicting NBS gene function:

• Random Forest and Gradient Boosting models for classifying NBS genes into functional categories based on sequence and structural features
• Recurrent Neural Networks (RNNs), particularly Long Short-Term Memory (LSTM) networks, for modeling expression dynamics over time following stress treatment
• Convolutional Neural Networks (CNNs) for identifying predictive cis-regulatory elements in promoter sequences
• Graph Neural Networks (GNNs) for leveraging protein-protein interaction networks and evolutionary relationships
• Multi-task Learning architectures that simultaneously predict multiple functional attributes (e.g., subcellular localization, stress responsiveness, pathogen specificity)

The exceptional diversity of NBS genes necessitates specialized approaches to address class imbalance issues, potentially through synthetic data generation techniques or specialized loss functions that weight minority classes more heavily.

Experimental Data for Model Training and Validation

Expression Profiling Under Stress Conditions

Table 3: Experimentally Validated Stress-Responsive NBS Genes as Training Data

Gene Name/Species	Stress Condition	Expression Response	Function Validated	Reference
SRC4 (Glycine max)	SMV infection, SA treatment, Ca2+ supplementation, temperature stress	Peak expression at 2-5 hpi; induced by all treatments; high basal expression	Antiviral activity; enhanced tolerance to 12°C and 37°C	[56]
GaNBS (Gossypium hirsutum)	Cotton leaf curl disease (CLCuD)	Upregulated in resistant accession	Virus tittering (validated by VIGS)	[4]
NBS genes (Asparagus officinalis)	Phomopsis asparagi infection	Majority unchanged or downregulated in susceptible cultivar	Potential functional impairment in domestication	[9]
OsUSP family (Oryza sativa)	Multiple abiotic stresses	24/46 significantly induced; LOCOs02g54590 & LOCOs05g37970 upregulated under all stresses	Stress adaptation mechanisms	[57]

Large-scale expression profiling studies provide critical datasets for training machine learning models to predict NBS gene function. A systematic analysis of 4085 soybean transcriptome datasets combined with SMV inoculation experiments revealed that SRC4 exhibited significantly higher basal expression than typical R genes and was induced by SMV infection, SA treatment, and Ca2+ supplementation, with peak expression at 2-5 hours post-treatment [56]. This precise kinetic information is invaluable for temporal function prediction.

In rice, expression profiling of Universal Stress Protein (USP) family genes identified 24 OsUSPs that were significantly induced under various stress conditions, with LOCOs02g54590 and LOCOs05g37970 emerging as particularly notable due to their broad-spectrum responsiveness, being upregulated under all tested stress conditions [57]. Such broad-spectrum responders represent valuable targets for both breeding applications and model validation.

Functional Validation Through Genetic Approaches

Several methodologies provide functional validation for NBS genes, creating gold-standard labels for supervised learning:

• Virus-Induced Gene Silencing (VIGS): Used to validate the role of GaNBS (OG2) in resistant cotton, demonstrating its putative role in virus tittering [4]
• Transgenic Approaches: ProSRC4::GUS reporter vectors in tobacco and transgenic Arabidopsis revealed that SRC4 transcriptional regulation is mediated through SA signaling pathways [56]
• Heterologous Expression: Maize NBS-LRR gene improved resistance to Pseudomonas syringae in Arabidopsis thaliana [28]
• Overexpression Studies: Transgenic plants overexpressing SRC4 exhibited enhanced tolerance to both 12°C and 37°C temperature stress [56]

These functional validation experiments not only confirm gene functions but also provide reliable labeled data for training machine learning models, with the experimental outcomes serving as ground truth for predictive algorithms.

Signaling Pathways as Predictive Features

Integrated Ca2+ and Salicylic Acid Signaling Network

The integration of signaling pathway information significantly enhances the predictive power of machine learning models for NBS gene function. Research has revealed that Ca2+ and salicylic acid (SA) serve as early signaling molecules and core defense hormones in plant immune responses, respectively, forming a highly integrated signaling cascade [56].

Figure 1: Integrated Ca²⁺ and SA Signaling Pathway Regulating NBS Gene Expression

This intricate signaling network involves several key components that can serve as predictive features in machine learning models:

• Calcium Signaling: When plants recognize pathogen-associated molecular patterns (PAMPs) or effector molecules, they rapidly activate plasma membrane and intracellular Ca2+ channels, leading to transient elevation of cytoplasmic Ca2+ concentrations [56]. These Ca2+ signals possess specific spatiotemporal patterns that can be precisely recognized and decoded by intracellular Ca2+-sensing proteins.

• Transcriptional Regulators: CBP60g serves as a key Ca2+-responsive transcription factor, sensing Ca2+ signal changes through its conserved calmodulin-binding domain [56]. In sard1 cbp60g double mutants, pathogen-induced ICS1 upregulation and SA accumulation are almost completely blocked, resulting in basal resistance defects and loss of systemic acquired resistance (SAR) [56].

• Negative Regulation: Calmodulin-binding transcriptional activator (CAMTA) family proteins serve as important negative regulatory factors, playing key roles in Ca2+ signal transduction [56]. CAMTA1, CAMTA2, and CAMTA3 negatively regulate SA biosynthesis by directly suppressing CBP60g and SARD1 gene expression.

Machine learning models can leverage the expression patterns of these signaling components as predictive features for NBS gene responsiveness, creating more accurate classifiers than those based solely on sequence characteristics.

Temperature Stress Integration in Immune Signaling

Temperature significantly influences NBS gene expression and function, providing additional predictive features for machine learning models. The soybean SRC4 gene demonstrates a dual role in both biotic and abiotic stress responses, particularly in temperature stress, with transgenic plants overexpressing SRC4 exhibiting enhanced tolerance to both 12°C and 37°C temperature stress [56].

Temperature changes can regulate the expression intensity and spatiotemporal patterns of R genes through multiple mechanisms [56]. Many NBS-LRR resistance genes exhibit upregulated expression at the transcriptional level under low-temperature conditions, which may represent an adaptive strategy for plants responding to increased pathogen invasion risks in low-temperature environments [56]. Conversely, high-temperature stress often suppresses the expression of certain R genes, leading to increased plant susceptibility to pathogens.

Table 4: Research Reagent Solutions for NBS Gene Functional Analysis

Reagent/Resource	Specific Examples	Application in NBS Gene Research	Reference
Genome Databases	Ensembl Plants, Phytozome, Plaza, NCBI	Genomic sequence retrieval, comparative analysis	[4] [57]
Domain Analysis Tools	HMMER, Pfam, CDD, SMART, InterProScan	NBS domain identification, classification	[28] [53]
Promoter Analysis Tools	PlantCARE, MEME Suite	Cis-element identification, motif discovery	[9] [53]
Expression Databases	IPF Database, CottonFGD, NCBI SRA	RNA-seq data retrieval, expression profiling	[4]
VIGS Vectors	TRV-based vectors, pTY vectors	Functional validation through gene silencing	[4]
Reporter Constructs	GUS, GFP, YFP fusion vectors	Promoter activity analysis, protein localization	[56]
Sequence Alignment Tools	Clustal Omega, MUSCLE, MAFFT	Phylogenetic analysis, conserved residue identification	[28] [53]
Phylogenetic Tools	MEGA, OrthoFinder, FastTree	Evolutionary analysis, orthogroup clustering	[9] [4]

This comprehensive toolkit enables researchers to generate the multi-modal data required for training effective machine learning models. The integration of data from these diverse resources addresses the challenge of limited labeled examples for specific NBS gene functions.

Machine learning approaches for predicting NBS gene function represent a paradigm shift in plant immunity research, moving from labor-intensive empirical studies to computationally-driven predictive science. The integration of diverse data types—from sequence features and expression profiles to evolutionary patterns and signaling network contexts—enables the development of models with remarkable predictive power.

Future advancements in this field will likely focus on several key areas:

• Multi-omics integration combining genomic, transcriptomic, epigenomic, and proteomic data
• Transfer learning approaches that leverage knowledge from well-characterized model species to predict gene function in less-studied crops
• Explainable AI methods that not only predict function but also identify the molecular basis for these predictions
• Single-cell genomics applications to understand cell-type-specific NBS gene functions
• Integration with protein structure prediction tools like AlphaFold to relate structural features to biological function

As these computational approaches mature, they will accelerate the identification of valuable NBS genes for crop improvement programs, potentially enabling the development of cultivars with enhanced resilience to the combined challenges of pathogen pressure and environmental stress. The unique dual functionality of certain NBS genes like SRC4 in both biotic and abiotic stress responses [56] highlights the potential for discovering multifunctional genetic elements that can address multiple agricultural constraints simultaneously.

Navigating Analytical Challenges in NBS Gene Family Studies

In comparative genomics, the identification and analysis of Nucleotide-Binding Site (NBS) domain genes are fundamental to understanding plant immune systems and disease resistance mechanisms [4] [58]. These genes, which constitute one of the largest resistance (R) gene families, encode proteins that recognize pathogen-derived molecules and initiate robust defense responses [59] [28]. The completeness of NBS gene identification is intrinsically linked to the quality of the underlying genome annotation, which is influenced by multiple factors including assembly contiguity, gene prediction algorithms, and supporting transcriptomic evidence [60] [61]. This guide provides a comparative analysis of genome annotation quality assessment tools and their measurable impact on the comprehensive characterization of NBS gene families, offering researchers a framework for selecting appropriate methodologies based on specific project requirements.

The Critical Link Between Annotation Quality and NBS Gene Research

Genome annotation quality directly determines the accuracy and completeness of downstream comparative genomic analyses. For NBS gene research in plants, incomplete or erroneous annotations can lead to significant underestimation of gene family sizes, misclassification of domain architectures, and flawed evolutionary inferences [4] [62]. The NBS-LRR gene family exhibits remarkable diversity in number and structure across plant species, with counts ranging from 73 in Akebia trifoliata to 2,151 in Triticum aestivum [28]. This variation reflects both biological differences and technical challenges in gene identification. Studies have demonstrated that annotation inconsistencies can substantially impact reported NBS gene counts; for example, different annotation approaches applied to the same Citrus sinensis genome have yielded varying inventories of NBS genes, affecting comparative analyses across citrus species [62].

The domain architecture of NBS genes further complicates accurate annotation. These genes are classified into multiple subfamilies—including CNL, TNL, NL, RNL, and others—based on their N-terminal domains (CC, TIR, or RPW8) and C-terminal LRR regions [4] [28]. Accurate identification requires precise delineation of these often-divergent domains, which may be fragmented in draft genomes or missed entirely by ab initio prediction tools [62]. The functional implications of incomplete NBS gene annotation are substantial, as these genes mediate resistance to diverse pathogens including viruses, bacteria, and fungi [59] [28]. In Nicotiana tabacum, for instance, comprehensive annotation revealed 603 NBS genes, with distinct distributions across architectural classes that provide insights into immune system evolution and potential disease resistance applications [28].

Comparative Analysis of Genome Annotation Quality Assessment Tools

Various computational frameworks have been developed to assess genome assembly and annotation quality, each employing distinct metrics and approaches. The table below compares four prominent tools used in contemporary genomics research.

Table 1: Comparison of Genome Annotation Quality Assessment Tools

Tool	Primary Methodology	Key Metrics	Strengths	Limitations
OMArk	Alignment-free protein comparisons to precomputed gene families [63]	Taxonomic consistency, completeness, contamination detection	Assesses both missing genes and spurious annotations; identifies contamination [63]	Requires proteome as input; overestimates completeness in high-duplication genomes [63]
BUSCO	Conservation-based universal single-copy orthologs [64]	Complete, duplicated, fragmented, and missing orthologs	Widely adopted; intuitive metrics; works on genome and transcriptome [64]	Limited to conserved gene space; blind to gene overprediction [63]
GenomeQC	Integrated metric calculation with benchmarking [64]	N50/NG50, L50/LG50, BUSCO scores, LTR Assembly Index (LAI)	Comprehensive assembly and annotation metrics; user-friendly web interface [64]	Primarily focused on assembly contiguity and completeness [64]
Annotation Consistency Tools	RNA-seq mapping and quantification statistics [60]	Mapping rates, transcript diversity, quantification success rates	Directly measures functional annotation utility for NGS applications [60]	Requires substantial RNA-seq data for assessment [60]

These tools collectively address different dimensions of annotation quality, from gene space completeness (BUSCO) to taxonomic consistency (OMArk) and assembly contiguity (GenomeQC). For NBS gene research, a combinatorial approach leveraging multiple assessment methods provides the most reliable evaluation of annotation suitability.

Experimental Approaches for Validating NBS Gene Annotations

Standardized NBS Gene Identification Pipeline

The accurate identification of NBS genes across multiple genomes requires a consistent bioinformatic workflow. The following methodology has been successfully applied in recent comparative studies of plant species:

Table 2: Key Research Reagent Solutions for NBS Gene Identification

Research Reagent	Function in NBS Gene Identification	Example Implementation
HMMER Suite	Hidden Markov Model-based domain detection [28] [62]	PF00931 (NB-ARC domain) search with e-value cutoff 1.1e-50 [4]
Pfam Domain Database	Confirmation of associated protein domains [4] [28]	Identification of TIR (PF01582), LRR (PF00560), and other accessory domains
NCBI Conserved Domain Database (CDD)	Validation of domain completeness and boundaries [28]	Verification of CC, TIR, and NBS domain architecture
OrthoFinder	Orthogroup inference and gene family evolution [4]	Clustering of NBS genes across multiple species
MCScanX	Detection of gene duplication events [28]	Identification of tandem and segmental duplications in NBS genes

The experimental workflow begins with domain identification using HMMER with the PF00931 (NB-ARC) model from Pfam, typically employing an e-value cutoff of 0.1 to 1.1e-50 to balance sensitivity and specificity [4] [62]. Candidate genes then undergo domain architecture characterization using Pfam and CDD to identify associated domains (TIR, CC, LRR). This is followed by phylogenetic analysis using tools such as MUSCLE for alignment and FastTree or MEGA for tree construction [28] [62]. Finally, evolutionary analyses investigate duplication patterns using MCScanX and selection pressures using KaKs_Calculator [28].

Diagram 1: NBS Gene Identification Workflow

Impact of Annotation Quality on NBS Gene Discovery: Case Studies

Several studies have directly demonstrated how annotation quality affects NBS gene identification. In a comparison of three Citrus genomes, researchers found that annotation methodology significantly influenced the reported number and diversity of NBS genes [62]. The study, which identified NBS genes in C. clementina, C. sinensis from the USA, and C. sinensis from China, revealed that variations in assembly quality and annotation approaches led to differing inventories of NBS genes, particularly affecting the identification of non-TIR types.

In Nicotiana species, a comprehensive analysis leveraging high-quality genome assemblies revealed 1,226 NBS genes across three species, with distinct distributions between diploid and tetraploid species [28]. The research demonstrated that whole-genome duplication events contributed significantly to NBS gene expansion, a finding that depended on contiguous assemblies and complete annotations to accurately resolve duplicated regions. The study further correlated annotation consistency with functional analysis, showing that improved assemblies enabled more reliable expression profiling of NBS genes in response to pathogens.

Best Practices for Annotation-Driven NBS Gene Analysis

Based on comparative assessments of annotation tools and their application to NBS gene research, the following practices are recommended for maximizing identification completeness:

• Implement Multi-Tool Quality Assessment: Combine BUSCO for completeness evaluation with OMArk for consistency checking and contamination detection [63] [64]. This approach provides complementary insights into different aspects of annotation quality that collectively impact NBS gene identification.

• Utilize Same-Species Transcriptomic Evidence: Incorporate RNA-seq data from the target species to improve gene model accuracy, particularly for defining UTRs and alternative splicing events [61]. Studies show that annotations incorporating same-species transcriptomic evidence yield more complete inventories of NBS genes and their variants [60].

• Apply Iterative Annotation Refinement: Use initial NBS gene identifications to guide targeted improvement of gene models, particularly for complex regions with tandem duplications [4] [28]. This iterative process helps resolve challenging genomic regions that may contain clustered NBS genes.

• Benchmark Against Curated Reference Sets: When available, compare identified NBS genes against manually curated reference sets from closely related species to assess identification efficiency and classify missing genes [63].

Diagram 2: Annotation Dependencies for NBS Gene Research

The completeness of NBS gene identification is fundamentally constrained by the quality of genome annotations. As demonstrated through comparative analyses of assessment tools and empirical studies across plant species, annotation quality directly impacts all aspects of NBS gene research—from basic inventories and classification to evolutionary and functional analyses. Researchers must prioritize annotation quality assessment as an integral component of comparative genomic studies of disease resistance genes, employing multiple complementary tools to evaluate different dimensions of quality. By adopting the standardized methodologies and best practices outlined in this guide, researchers can significantly improve the reliability and biological relevance of their NBS gene analyses, ultimately advancing our understanding of plant immune systems and enabling more effective strategies for crop improvement.

The nucleotide-binding site (NBS) domain is a critical component of the largest class of plant disease resistance (R) genes, which encode proteins that recognize diverse pathogens and initiate robust immune responses [65] [66]. In the field of comparative genomics, accurately distinguishing functionally intact NBS-encoding genes from non-functional pseudogenes is a fundamental challenge with significant implications for disease resistance breeding and evolutionary studies [67]. Pseudogenes—non-functional genomic sequences resembling functional genes—arise from duplicated genes that accumulate disabling mutations, such as premature stop codons and frameshift mutations, rendering them unable to produce functional proteins [67].

Domain integrity assessment provides the methodological foundation for this discrimination, leveraging the characteristic domain architecture of NBS-encoding resistance genes. This guide systematically compares experimental approaches for evaluating NBS domain integrity across plant species, providing researchers with standardized protocols and analytical frameworks to advance functional genomics in plant immunity.

Structural Organization of NBS Domain Genes

Conserved Domain Architecture

NBS-encoding resistance genes typically encode proteins containing a conserved nucleotide-binding site (NBS) domain and often additional domains that define their functional classification [68] [66]. The general structural organization includes:

• N-terminal domain: Typically a Toll/interleukin-1 receptor (TIR) domain, coiled-coil (CC) domain, or resistance to powdery mildew 8 (RPW8) domain [68] [69]
• Central NBS domain: Contains several highly conserved motifs in strict order [65] [67]
• C-terminal domain: Often composed of leucine-rich repeats (LRR) that facilitate protein-protein interactions and pathogen recognition [70] [66]

Based on their N-terminal domains, NBS-LRR genes are classified into three major subfamilies: TNL (TIR-NBS-LRR), CNL (CC-NBS-LRR), and RNL (RPW8-NBS-LRR) [68] [69]. Additionally, many atypical configurations exist where one or more domains are absent (e.g., NBS-only, TN, CN, NL) [66].

Conserved Motifs Within the NBS Domain

The NBS domain itself contains several conserved motifs that maintain strict order across plant species. Motif analysis across Triticeae species confirmed the presence of six commonly conserved motifs: P-loop, RNBS-A, Kinase-2, Kinase-3a, RNBS-C, and GLPL [65]. Research across 34 plant species revealed 168 distinct domain architecture patterns, encompassing both classical configurations and species-specific structural variations [4].

Table 1: Conserved Motifs in the NBS Domain

Motif Name	Conserved Sequence	Functional Role
P-loop	GKTT/T	ATP/GTP binding
RNBS-A	FLHIACF	Structural stability
Kinase-2	LVLDDVW	Hydrolytic activity
Kinase-3a	GSRIIITTRD	Signal transduction
RNBS-C	CFLYCALFPL	Unknown
GLPL	GMGLPLA	Structural motif

Methodological Framework for Domain Integrity Assessment

Computational Identification and Domain Annotation

The initial step in domain integrity assessment involves comprehensive identification of NBS-encoding sequences within plant genomes using integrated computational approaches:

Figure 1: Computational workflow for identifying and classifying NBS-encoding genes

Hidden Markov Model (HMM) Searches

The most reliable method for initial identification involves using HMMER software with the NB-ARC domain (PF00931) HMM profile from the Pfam database [68] [71] [67]. The standard protocol includes:

• Database Preparation: Compile predicted protein sequences from the target genome
• HMMER Scanning: Execute hmmsearch with the NB-ARC domain profile (E-value threshold typically <1.0) [68]
• Candidate Extraction: Extract all sequences exceeding significance thresholds

As demonstrated in Akebia trifoliata research, this approach identified 73 NBS genes when combined with additional validation steps [68].

Domain Annotation and Validation

Following identification, candidate sequences require comprehensive domain annotation using integrated tools:

• NCBI Conserved Domain Database (CDD): Verify presence of NBS and associated domains [68]
• Pfam HMM Searches: Identify TIR (PF01582) and LRR (PF08191) domains [68] [67]
• Coiled-coil Prediction: Use MARCOIL or COILS with threshold probability of 90% [67]
• MEME Suite: Identify conserved motifs within the NBS domain [67]

In the Solanum tuberosum study, researchers developed a species-specific NBS HMM model to improve identification accuracy [67].

Criteria for Distinguishing Functional Genes from Pseudogenes

The critical assessment of domain integrity focuses on identifying disruptive mutations that compromise protein function:

Table 2: Diagnostic Features for Discriminating Functional Genes from Pseudogenes

Feature	Functional Gene	Pseudogene
Open Reading Frame	Complete, uninterrupted	Premature stop codons, frameshifts
Conserved motifs	All motifs present and intact	Missing or truncated motifs
Domain architecture	Complete domains	Partial or missing domains
Transcript evidence	Expression supported by RNA-seq	No expression evidence
Selective pressure	Ka/Ks < 1 (purifying selection)	Ka/Ks ≈ 1 (neutral evolution)

Assessment of Disabling Mutations

Pseudogenes typically contain disabling mutations that disrupt the reading frame or introduce premature termination:

• Premature Stop Codons: Truncate the protein before complete domain assembly [67]
• Frameshift Mutations: Disrupt the reading frame, altering downstream sequences [67]
• Splice Site Mutations: Affect proper mRNA processing and domain integrity
• Partial Domain Deletions: Remove critical functional regions

In Solanum tuberosum, approximately 41% (179 of 435) of NBS-encoding genes were classified as pseudogenes, primarily due to premature stop codons and frameshift mutations [67].

Structural Integrity Evaluation

Functional NBS-encoding genes must maintain structural integrity across several dimensions:

• Complete NBS Domain: All conserved motifs (P-loop to GLPL) must be present and intact
• Intact Flanking Domains: TIR/CC at N-terminus and LRR at C-terminus when present
• Conserved Residues: Critical amino acids for nucleotide binding must be preserved

Research in Vernicia species demonstrated that susceptible V. fordii lacked certain LRR domains present in resistant V. montana, highlighting the functional importance of domain completeness [70].

Comparative Genomics of NBS Genes Across Species

Variation in NBS Gene Family Size and Composition

Comparative analysis across diverse plant species reveals substantial variation in NBS gene family size and composition:

Table 3: Comparative Analysis of NBS-Encoding Genes Across Plant Species

Plant Species	Family/Group	Total NBS Genes	Functional	Pseudogenes	Notable Features
Solanum tuberosum (potato)	Solanaceae	435	256	179 (41%)	High pseudogene percentage
Akebia trifoliata	Lardizabalaceae	73	73	Not reported	50 CNL, 19 TNL, 4 RNL
Salvia miltiorrhiza	Lamiaceae	196	62 complete	Not reported	61 CNL, 1 RNL, no TNL
Vernicia montana	Euphorbiaceae	149	149	Not reported	9 CC-NBS-LRR, 3 TIR-NBS-LRR
Vernicia fordii	Euphorbiaceae	90	90	Not reported	No TIR domains
Ipomoea batatas (sweet potato)	Convolvulaceae	889	Not reported	Not reported	Highest count among Ipomoea
Grass pea (Lathyrus sativus)	Fabaceae	274	274	Not reported	124 TNL, 150 CNL
Arabidopsis thaliana	Brassicaceae	207	167	Not reported	Model for eudicot NBS genes

Evolutionary Patterns Affecting Domain Integrity

The evolutionary dynamics of NBS genes significantly impact their functional status:

Gene Duplication Mechanisms

• Tandem Duplications: Lead to gene clusters with related specificities [68] [69]
• Segmental Duplications: Copy large chromosomal regions containing multiple genes [4]
• Whole Genome Duplication: Polyploidization events creating multiple gene copies [4]

In Akebia trifoliata, tandem and dispersed duplications produced 33 and 29 NBS genes respectively, representing the main forces for NBS gene expansion [68].

Birth-and-Death Evolution

NBS gene families evolve primarily through a birth-and-death process where:

• New genes are created by duplication
• Some duplicates acquire new functions
• Others accumulate mutations and become pseudogenes
• Non-functional copies are eventually eliminated [22]

This evolutionary pattern creates genomic landscapes where functional genes and pseudogenes coexist in complex arrangements.

Experimental Validation of Functional Status

Transcriptomic Analysis

RNA sequencing provides critical evidence for functional gene status by verifying expression:

• Tissue-Specific Expression: Functional genes show regulated expression across tissues
• Induction Under Stress: Genuine R genes are often upregulated during pathogen challenge
• Alternative Splicing: Complex transcription patterns indicate functional regulation

In Salvia miltiorrhiza, expression profiling of SmNBS-LRR genes revealed close association with secondary metabolism and stress responses [66]. Similarly, transcriptome analysis of resistant and susceptible sweet potato cultivars identified differentially expressed NBS genes responding to stem nematodes and Ceratocystis fimbriata infection [69].

Functional Characterization Approaches

Virus-Induced Gene Silencing (VIGS)

VIGS provides direct evidence for gene function by knocking down candidate genes and assessing phenotypic consequences:

• In Vernicia montana, VIGS of VmNBS-LRR demonstrated its essential role in Fusarium wilt resistance [70]
• In cotton, silencing of GaNBS (OG2) increased susceptibility to cotton leaf curl disease [4]

Quantitative PCR Validation

Targeted qPCR analysis confirms expression patterns suggested by RNA-seq:

• In grass pea, nine LsNBS genes were analyzed under salt stress conditions, showing differential expression patterns [71]
• In sweet potato, six differentially expressed NBS genes were validated by qRT-PCR with results consistent with transcriptome data [69]

Table 4: Essential Research Reagents for NBS Gene Analysis

Reagent/Resource	Specific Examples	Application	Key Features
HMM Profiles	Pfam NB-ARC (PF00931), TIR (PF01582), LRR (PF08191)	Domain identification	Curated protein family models
Genomic Resources	NCBI Genome, Phytozome, Plaza	Comparative genomics	Multi-species genomic data
Software Tools	HMMER, MEME, NCBI CDD, MARCOIL	Domain analysis	Specialized algorithms
Expression Databases	IPF Database, CottonFGD, NCBI BioProject	Transcriptomic validation	Tissue/stress-specific data
PCR Reagents	Degenerate primers for NBS motifs	Gene isolation	Target conserved motifs

• Bioinformatic Pipelines: OrthoFinder for orthogroup analysis, DIAMOND for sequence similarity searches, and MAFFT for multiple sequence alignment facilitate comparative genomics [4]
• Experimental Validation Tools: Quantitative RT-PCR systems, VIGS vectors, and recombinant protein expression systems enable functional characterization [70] [71]

Domain integrity assessment provides a powerful framework for distinguishing functional NBS genes from pseudogenes, combining computational prediction with experimental validation. The conserved architecture of NBS domains enables systematic evaluation across plant species, revealing diverse evolutionary trajectories including gene family expansions, contractions, and frequent pseudogenization. As genomic resources continue to expand, integrated approaches that leverage both comparative genomics and functional characterization will be essential for unlocking the potential of NBS genes in crop improvement and sustainable agriculture.

The accurate resolution of tandem duplication complexes represents a fundamental challenge in comparative genomics, particularly in the study of rapidly evolving gene families such as plant nucleotide-binding site (NBS) domain genes. Tandem duplication, characterized by the adjacent repetition of genomic regions, serves as a primary mechanism for gene family expansion and functional diversification in eukaryotes [72] [73]. In plant genomes, this process has generated extensive arrays of NBS-encoding genes that play crucial roles in pathogen recognition and disease resistance [4] [14]. The inherent complexity of these regions—marked by high sequence similarity, structural variation, and dynamic evolutionary histories—complicates precise gene annotation and enumeration.

Resolving these complexes is not merely a technical exercise but a prerequisite for understanding genome evolution and functional adaptation. Studies across plant species have revealed that tandem duplication contributes significantly to the species-specific amplification of NBS-encoding genes following whole genome triplication events [14]. For instance, in Brassica species, tandem duplicates have been selectively maintained and exhibit differential expression patterns, suggesting their importance in adaptive evolution [14]. The strategic resolution of these regions enables researchers to accurately reconstruct evolutionary histories, identify candidate genes for disease resistance, and decipher the molecular arms races between plants and their pathogens [4] [74].

Methodological Approaches for Resolving Tandem Duplications

Computational Detection and Annotation Tools

Multiple bioinformatic approaches have been developed to detect tandem duplications, each with distinct strengths, limitations, and optimal use cases. The selection of an appropriate method depends heavily on the evolutionary age of the duplication, the genomic context, and the specific research questions being addressed.

Table 1: Comparative Analysis of Computational Tools for Tandem Duplication Detection

Tool Name	Primary Methodology	Optimal Use Case	Strengths	Limitations
ReD Tandem	Flow-based chaining of DNA-level self-alignment anchors [75]	Agnostic identification of recent tandem duplications without annotation dependency	Detects non-coding duplicates (pseudogenes, RNA genes); complements protein-based methods [75]	Inherently restricted to relatively recent duplications [75]
OrthoFinder	DIAMOND for sequence similarity; MCL clustering algorithm [4]	Evolutionary orthogroup analysis across multiple species	Identifies core and species-specific orthogroups; integrates with phylogenetic analysis [4]	Relies on annotated gene models; may miss non-coding elements
HMMER	Hidden Markov Models with Pfam domain profiles (e.g., NBS domain PF00931) [14]	Family-specific identification of domain-encoding genes	High accuracy for identifying genes with specific conserved domains; uses trusted cutoffs [14]	Limited to known domain architectures; may miss divergent copies
SynNet	Synteny network analysis [76]	Studying genomic arrangements of protein-coding genes in plants	Reveals evolutionary relationships through synteny conservation [76]	Requires multiple genome sequences for comparative analysis

Experimental Validation Techniques

Computational predictions require experimental validation to confirm both the physical presence and functional implications of tandem duplications. Several laboratory techniques provide this essential verification.

Microarray-based Comparative Genomic Hybridization (CGH) offers a robust method for initial duplication screening across related species. The experimental workflow involves digesting genomic DNA with DNaseI, labeling the 3' termini of fragmentation products with biotin-dideoxyuridine triphosphate (ddUTP), and hybridizing the target fragments onto platform-specific arrays (e.g., Affymetrix GeneChip). The resulting hybridization intensity ratios between species are calculated for each probe, with median fold-change values serving as thresholds for duplication criteria [72]. This approach successfully identified a three-gene cluster in Drosophila created by two rounds of tandem duplication within a 5-million-year timeframe [72].

Whole-Genome Sequencing (WGS) coupled with structural variation analysis provides nucleotide-level resolution of tandem duplication events. The standard protocol involves sequencing genomic DNA to sufficient coverage (typically 30x or higher), aligning reads to a reference genome, and applying specialized algorithms to detect duplication signatures. In a comprehensive study of gastric cancer genomes, researchers analyzed 168 whole genomes to identify tandem duplication hotspots, validating predictions through PCR and Sanger sequencing (achieving 95% validation rate on tested candidates) [77]. This approach revealed diverse models of complex structural variations leading to oncogene amplification through tandem duplications.

Expression Profiling determines the functional consequences of tandem duplications through transcriptomic analysis. RNA sequencing (RNA-seq) from multiple tissues and developmental stages, or under various stress conditions, can reveal expression divergence among tandem duplicates. Standard methodology includes total RNA extraction (e.g., using Qiagen kits), library preparation, sequencing, and quantification of expression values (e.g., FPKM - Fragments Per Kilobase of transcript per Million mapped reads). Studies in cotton have demonstrated that NBS-encoding genes in specific orthogroups (OG2, OG6, OG15) show upregulated expression in response to biotic and abiotic stresses, suggesting functional specialization of tandem duplicates [4].

Table 2: Experimental Methods for Validating Tandem Duplications

Method	Key Reagents/Equipment	Primary Output	Resolution	Throughput
CGH	DNaseI, biotin-ddUTP, microarray platform, hybridization equipment [72]	Hybridization intensity ratios indicating copy number variation [72]	Gene-level	Medium
WGS	High-throughput sequencer, PCR reagents, Sanger sequencer for validation [77]	Comprehensive structural variant catalog including tandem duplications [77]	Nucleotide-level	High
RNA-seq	RNA extraction kits, library preparation reagents, sequencing platform [4]	Expression profiles (FPKM) across tissues and conditions [4]	Transcript-level	High
VIGS	Agrobacterium strains, silencing vectors, plant inoculation supplies [4]	Functional validation through phenotypic assessment of silenced genes [4]	Gene-level	Low

Experimental Data and Case Studies

Plant NBS Domain Gene Families

Comprehensive genomic analysis across 34 plant species identified 12,820 NBS-domain-containing genes classified into 168 distinct domain architecture classes [4]. This study revealed remarkable diversification beyond classical structures (NBS, NBS-LRR, TIR-NBS, TIR-NBS-LRR) to include species-specific patterns such as TIR-NBS-TIR-Cupin1-Cupin1 and Sugar_tr-NBS. Orthogroup analysis delineated 603 orthogroups, with both core (widely conserved) and unique (species-specific) groups showing significant expansion through tandem duplication [4].

In Brassica species, comparative analysis with Arabidopsis thaliana revealed distinct evolutionary trajectories following whole genome triplication. Researchers identified 157 and 206 NBS-encoding genes in B. oleracea and B. rapa genomes, respectively [14]. Phylogenetic analysis classified these into six subgroups, with tandem duplication driving species-specific amplification after the divergence of B. rapa and B. oleracea. Expression profiling of orthologous gene pairs demonstrated differential expression patterns between the two species, suggesting subfunctionalization or neofunctionalization of tandem duplicates [14].

Functional Divergence in Drosophila

Molecular population genetic analysis of a three-gene cluster in Drosophila melanogaster (CG32708, CG32706, and CG6999) revealed how tandem duplicates acquire novel functions. This cluster originated through two rounds of tandem duplication within the last 5 million years, with CG32708 as the parental copy, CG32706 originating in the ancestor of Drosophila simulans and D. melanogaster, and CG6999 being the newest duplicate unique to D. melanogaster [72]. Despite sequence similarity, all three genes exhibited divergent expression profiles, with CG6999 acquiring a novel transcript. Population genetic tests, including McDonald-Kreitman analysis, provided evidence that the evolution of CG6999 and CG32706 was driven by positive Darwinian selection [72].

Coevolutionary Arms Races

The SERPINA gene family in rodents exemplifies how tandem duplication fuels coevolutionary arms races between predators and prey. Genomic analysis revealed rapid birth-death evolution of SERPINA1-like and SERPINA3-like genes within and between rodent lineages [74]. In the Big-eared woodrat (Neotoma macrotis), which exhibits remarkable resistance to snake venom, researchers identified 12 paralogous duplicates of SERPINA3. Functional characterization demonstrated that two paralogs inhibited venom serine proteases, with one exhibiting neofunctionalization to inhibit both chymotrypsin-like and trypsin-like proteases simultaneously [74]. This exemplifies how tandem duplication generates functional diversity in response to selective pressures.

Experimental Protocols for Key Analyses

Genome-Wide Identification of NBS-Encoding Genes

Step 1: Domain Identification

• Retrieve Pfam HMM profiles for NBS (NB-ARC) domain (PF00931) and associated domains
• Perform HMM search against proteome using HMMER v3.0+ with "trusted cutoff" thresholds
• Curate initial candidate set and construct species-specific NBS profile using "hmmbuild"
• Conduct final search with refined model to identify high-confidence NBS-encoding genes [14]

Step 2: Domain Architecture Classification

• Identify N-terminal and C-terminal domains using HMMPfam and HMMSmart
• Confirm coiled-coil (CC) motifs using PAIRCOIL2 (P-score cut-off 0.025) and MARCOIL (threshold probability 90)
• Classify genes into structural categories (TNL, CNL, RNL, etc.) based on domain combinations [4] [14]

Step 3: Tandem Duplication Detection

• Map genes to chromosomes and identify clusters (≤10 intervening genes between paralogs)
• Calculate synonymous substitution rates (dS) to estimate duplication ages
• Perform phylogenetic analysis to validate evolutionary relationships [78]

Population Genetic Analysis for Selection Detection

Step 1: Polymorphism Data Collection

• Sequence target genes from multiple individuals/populations (20+ recommended)
• Extract and align sequences; identify polymorphic sites
• Calculate diversity indices (π, θ) using DnaSP or similar software [72]

Step 2: Neutrality Tests

• Perform Tajima's D, Fu & Li's, and Fay & Wu's tests
• Assess significance using coalescent simulations (2000+ replicates)
• Conduct McDonald-Kreitman test comparing polymorphism and divergence [72]

Step 3: Selection Inference

• Interpret significant deviations from neutrality
• For positive selection: significant excess of nonsynonymous substitutions in MK test
• For balancing selection: significantly positive Tajima's D and high diversity [72]

Visualization of Analytical Workflows

ReD Tandem Computational Pipeline

Diagram 1: ReD Tandem computational workflow for agnostic tandem duplication detection

Integrated Experimental-Computational Validation

Diagram 2: Integrated approach for tandem duplication complex validation

Table 3: Essential Research Reagents and Resources for Tandem Duplication Studies

Category	Specific Reagents/Resources	Function/Application	Example Use Case
Bioinformatics Tools	ReD Tandem [75], OrthoFinder [4], HMMER [14], DnaSP [72]	Detection, classification, and evolutionary analysis of duplicated genes	Identifying tandem arrays directly from genomic sequence [75]
Domain Databases	Pfam (NBS: PF00931, TIR: PF01582) [14]	Curated domain models for gene family identification	Classifying NBS-encoding genes into structural categories [14]
Genomic Resources	BRAD database [14], Bolbase [14], Phytozome [4], TAIR [14]	Annotated genome sequences and comparative genomics platforms	Comparative analysis of NBS genes across Brassica species [14]
Laboratory Reagents	DNaseI, biotin-ddUTP [72], Qiagen DNA/RNA extraction kits [72], Taq polymerase [72]	Nucleic acid preparation and manipulation for experimental validation	Microarray-based CGH for duplication detection [72]
Sequencing Platforms	Illumina for WGS [77], Applied Biosystems DNA sequencers [72]	High-throughput sequencing for structural variant detection	Identifying TD hotspots in gastric cancer genomes [77]
Functional Validation Tools	VIGS vectors [4], Agrobacterium strains [4], recombinant protein expression systems [74]	Assessing functional consequences of duplicated genes	Testing role of GaNBS (OG2) in virus resistance [4]

The resolution of tandem duplication complexes requires integrated methodological approaches that combine sophisticated computational detection with rigorous experimental validation. As genomic technologies advance, the research community will benefit from standardized protocols, improved algorithms for detecting ancient duplications, and enhanced functional characterization methods. The strategic resolution of these complex genomic regions continues to provide fundamental insights into genome evolution, adaptation mechanisms, and the molecular basis of disease resistance across diverse species.

Balancing Stringency and Sensitivity in Domain Detection Thresholds

In comparative genomics, the accurate identification of conserved protein domains forms the foundation for understanding gene family evolution and function. This is particularly critical for nucleotide-binding site (NBS) domain genes, which constitute one of the largest and most variable resistance gene families in plants [4]. The detection of these domains governs all downstream analyses, from gene family characterization to functional predictions. However, researchers face a fundamental methodological challenge: how to balance stringency and sensitivity in domain detection thresholds. Overly stringent thresholds risk excluding legitimate family members, while overly sensitive parameters may introduce false positives, compromising data integrity. This guide objectively compares the performance of different domain detection methodologies applied to NBS domain genes across plant species, providing experimental data to inform selection criteria for genomics researchers.

Methodological Approaches for Domain Detection

Hidden Markov Model (HMM)-Based Detection

HMM-based approaches represent the gold standard for domain identification, using probabilistic models built from multiple sequence alignments of known domains.

• Typical Experimental Protocol: The standard workflow begins with retrieving the NB-ARC domain (Pfam: PF00931) HMM profile. Researchers then perform HMM searches against target protein datasets using tools like HMMER v3.1b2, typically applying an E-value cutoff of 1e-5 to 1e-10 [9] [28]. Following initial identification, additional domains (TIR, CC, LRR) are characterized using InterProScan or NCBI's Conserved Domain Database (CDD) to classify NBS genes into subfamilies (CNL, TNL, RNL, etc.) [28].

• Performance Considerations: This method provides excellent reproducibility but requires careful threshold selection. Studies on Nicotiana species successfully identified 1,226 NBS genes across three genomes using this approach, demonstrating its comprehensive coverage [28].

Deep Learning-Based Prediction

Novel deep learning tools have emerged that bypass traditional domain detection, instead predicting resistance genes directly from protein sequences.

• PRGminer Workflow: This tool implements a two-phase prediction system: Phase I classifies input protein sequences as resistance genes or non-resistance genes, while Phase II categorizes predicted R-genes into eight structural classes (CNL, KIN, RLP, LECRK, RLK, LYK, TIR, TNL) [47] [79].

• Performance Metrics: PRGminer achieves impressive accuracy metrics, with 98.75% accuracy in k-fold testing and 95.72% on independent testing in Phase I, and 97.55% and 97.21% respectively in Phase II classification [79]. This represents a significant advancement over traditional methods, particularly for fragmented genes or those with low sequence homology.

Genome-Wide Comparison Frameworks

Large-scale comparative studies require standardized pipelines to ensure consistent domain detection across multiple species.

• OrthoFinder Analysis: This approach enables evolutionary comparison through orthogroup clustering, using DIAMOND for fast sequence similarity searches and the MCL clustering algorithm [4]. The methodology is particularly valuable for tracking NBS gene family expansion and contraction across evolutionary lineages.

• Cross-Species Validation: One study identified 12,820 NBS-domain-containing genes across 34 plant species, classifying them into 168 distinct domain architecture classes [4]. This large-scale analysis provides a critical reference dataset for validating domain detection thresholds.

Table 1: Domain Detection Methods and Performance Characteristics

Method	Key Tools	Strengths	Optimal E-value/Threshold	Representative Applications
HMM-Based	HMMER, InterProScan, CDD	High specificity, standardized parameters	E-value 1e-5 to 1e-10 [9] [28]	Nicotiana NBS census (1,226 genes) [28]
Deep Learning	PRGminer	Handles low-homology sequences, high accuracy	Classification accuracy 95.72-98.75% [79]	Plant resistance gene prediction across species
Comparative Genomics	OrthoFinder, MCScanX	Evolutionary context, orthology resolution	E-value 1e-10 for synteny [4]	12,820 NBS genes across 34 species [4]

Comparative Performance Across Plant Lineages

Detection Sensitivity and Taxonomic Range

The stringency of domain detection parameters directly impacts reported gene counts and evolutionary inferences. Studies employing consistent HMM thresholds have revealed remarkable variation in NBS gene abundance across plant taxa, from just 2 NLRs in Selaginella moellendorffii to over 2,000 in Triticum aestivum [4]. This variation reflects both biological reality and methodological sensitivity.

Critical findings include the complete absence of TNL genes in Poaceae family and the dicot Mimulus guttatus, discovered through systematic domain profiling [80]. Such lineage-specific losses would remain undetected with insufficiently sensitive detection parameters. Similarly, research on Asparagus species revealed NLR contraction from 63 genes in wild A. setaceus to just 27 in domesticated A. officinalis, with important implications for disease susceptibility [9].

Impact on Gene Classification and Annotation

Domain detection thresholds directly influence subsequent gene classification and functional prediction. In cowpea, comprehensive genome analysis identified 2,188 R-genes distributed across 29 classes, with kinases (KIN) and transmembrane proteins (RLKs and RLPs) predominating [19]. The accurate discrimination between these classes depends entirely on initial domain detection sensitivity.

Table 2: NBS Gene Distribution Across Plant Species Using Standardized Detection Methods

Plant Species	Total NBS Genes	CNL/CN	TNL/TN	Other/Partial	Detection Method
Nicotiana tabacum [28]	603	224 (37.1%)	73 (12.1%)	306 (50.8%)	HMM (PF00931) + CDD
Nicotiana sylvestris [28]	344	130 (37.8%)	42 (12.2%)	172 (50.0%)	HMM (PF00931) + CDD
Nicotiana tomentosiformis [28]	279	112 (40.1%)	40 (14.3%)	127 (45.5%)	HMM (PF00931) + CDD
Vigna unguiculata (cowpea) [19]	2,188 R-genes	Not specified	Not specified	29 classes total	HMM + manual curation
Asparagus setaceus [9]	63 NLRs	Not specified	Not specified	Not specified	HMM + BLASTp
Asparagus officinalis [9]	27 NLRs	Not specified	Not specified	Not specified	HMM + BLASTp

Experimental Protocols for Method Validation

Analytical Validation Framework

Rigorous validation of domain detection methods requires systematic experimental design. The BabyDetect study provides a exemplary model, implementing strict quality control thresholds for sequencing, coverage, and contamination across more than 5,900 samples [81]. Their workflow employed:

• Longitudinal Performance Monitoring: Tracking consistency across processing batches
• Automation Integration: Implementing automated DNA extraction to improve scalability
• Panel Redesign: Iteratively refining target regions to enhance coverage
• False Positive Mitigation: Focusing on known pathogenic/likely pathogenic variants to maintain clinical actionability [81]

Orthology-Based Benchmarking

Evolutionary validation through orthology analysis provides a critical method for verifying domain detection accuracy. One comprehensive study organized NBS genes into 603 orthogroups, identifying both core (widely conserved) and unique (lineage-specific) groups [4]. Expression profiling confirmed the functional relevance of these groups, with orthogroups OG2, OG6, and OG15 showing upregulated expression under biotic and abiotic stresses in cotton accessions with varying resistance to cotton leaf curl disease [4].

Domain Detection Workflow and Threshold Selection

Research Reagent Solutions Toolkit

Table 3: Essential Research Reagents and Computational Tools for NBS Domain Detection

Tool/Reagent	Specific Application	Function in Domain Detection	Example Implementation
HMMER Suite	HMM-based domain search	Identifies conserved domains using probabilistic models	NBS identification in Nicotiana (PF00931) [28]
Pfam Database	Domain profile repository	Provides curated HMM profiles for domain families	NB-ARC domain (PF00931) reference [28]
InterProScan	Integrated domain annotation	Combines multiple databases for comprehensive domain analysis	Domain architecture characterization [9]
CDD (NCBI)	Conserved domain identification	Annotates functional domains in protein sequences	Verification of CC, TIR, LRR domains [28]
PRGminer	Deep learning prediction	Classifies R-genes without direct domain detection	Alternative to HMM for low-homology sequences [47]
OrthoFinder	Orthogroup inference	Groups genes into orthologous groups across species	Evolutionary analysis of NBS genes [4]
MEME Suite	Motif discovery	Identifies conserved motifs within protein families	NBS domain motif analysis [9]

Method Selection Guide for Domain Detection

The balance between stringency and sensitivity in domain detection thresholds remains context-dependent, requiring researchers to align methodological choices with specific research objectives. For comprehensive gene family censuses, more sensitive HMM thresholds (E-value 1e-5) combined with manual curation provide optimal coverage. For evolutionary studies seeking orthologous relationships, intermediate stringency (E-value 1e-10) with orthology resolution offers the best balance. For non-model organisms or fragmented genomes, deep learning approaches like PRGminer circumvent limitations of traditional domain detection altogether. Critically, methodological transparency and threshold reporting enable meaningful comparisons across studies and species, advancing our understanding of NBS gene family evolution and function across the plant kingdom.

Integrating Transcriptomic Data to Filter Constitutively Expressed NBS Genes

The Nucleotide-Binding Site-Leucine-Rich Repeat (NBS-LRR) gene family represents one of the most important classes of plant disease resistance (R) genes, playing a critical role in effector-triggered immunity (ETI) by recognizing pathogen effector proteins and activating defense responses [28] [4]. As a library, NLM provides access to scientific literature. Inclusion in an NLM database does not imply endorsement of, or agreement with, the contents by NLM or the National Institutes of Health. Learn more: PMC Disclaimer | PMC Copyright Notice. Recent advances in comparative genomics have revealed remarkable diversity in NBS-LRR genes across plant species, with significant variation in gene number, structural architecture, and evolutionary patterns [4] [82]. The integration of transcriptomic data provides a powerful approach to filter constitutively expressed NBS genes, enabling researchers to identify core components of plant immune systems with consistent expression patterns across different conditions, tissues, and species. This guide objectively compares methodologies and resources for identifying and analyzing constitutively expressed NBS genes, providing experimental protocols and data frameworks for researchers in plant genomics and disease resistance breeding.

Comparative Analysis of NBS Gene Identification Pipelines

Genome-Wide Identification Methods

The accurate identification of NBS-LRR genes across plant genomes requires integrated bioinformatics approaches combining multiple detection methods. Table 1 compares the primary computational pipelines used in recent studies for genome-wide NBS gene identification.

Table 1: Comparison of NBS Gene Identification Methods and Tools

Method Category	Specific Tools	Key Parameters	Target Domain	Representative Studies
HMMER Search	HMMER v3.1b2	E-value threshold, PF00931 (NB-ARC) model	NBS domain	Nicotiana species (2025) [28]
Pfam Domain Analysis	PfamScan.pl	E-value (1.1e-50), Pfam-A_hmm model	Multiple domains	34 species analysis (2024) [4]
Conserved Domain Database	NCBI CDD	Default parameters, domain validation	TIR, CC, LRR domains	Nicotiana, Rosaceae studies [28] [82]
BLAST Search	BLASTP	E-value threshold (1.0), custom databases	Full-length sequences	Rosaceae species (2022) [82]

The integration of these complementary methods ensures comprehensive identification of NBS genes. The HMMER approach using the PF00931 model provides high sensitivity for detecting the conserved NB-ARC domain, while CDD and Pfam analyses enable accurate classification based on additional domains [28]. BLAST searches serve as a valuable supplementary method for identifying potential family members that may have divergent domain architectures.

Classification Systems for NBS Gene Families

NBS-LRR genes are classified based on their N-terminal domains and overall domain architecture. Table 2 presents the classification schemes and their distribution across recent multi-species studies.

Table 2: NBS Gene Classification Systems and Distribution Patterns

Classification System	Gene Categories	Domain Architecture	Species Examples	Percentage Distribution
Eight-Subfamily System [28]	CN, CNL, N, NL, RN, RNL, TN, TNL	Based on N-terminal and C-terminal domains	Nicotiana tabacum	NBS-only: 45.5%, CC-NBS: 23.3% [28]
Three-Subfamily System [82]	TNL, CNL, RNL	TIR/CC/RPW8-NBS-LRR	Rosaceae species	Varies by species [82]
Simplified Two-Subfamily [28]	TNL, non-TNL	Presence/absence of TIR domain	Solanaceae species	Dependent on evolutionary history
Domain Architecture Classes [4]	168 classes identified	Classical and species-specific patterns	34 plant species	Includes novel domain combinations

The classification approach significantly impacts the interpretation of evolutionary patterns and functional characterization. Studies on Nicotiana species revealed that approximately 45.5% of NBS genes contain only the NBS domain without LRR regions, followed by CC-NBS types at 23.3%, while TIR-NBS members were the least abundant [28]. This distribution varies substantially across plant families, reflecting species-specific evolutionary trajectories.

Transcriptomic Integration for Constitutive Expression Analysis

Experimental Design for Transcriptome Studies

The identification of constitutively expressed NBS genes requires carefully designed transcriptomic experiments that capture expression patterns across multiple conditions, tissues, and developmental stages. Key considerations include:

• Temporal Sampling: Studies in banana blood disease resistance collected root tissue samples at 12 hours, 1 day, and 7 days post-inoculation to capture early and late response patterns [83].
• Spatial Sampling: Research on cotton NBS genes analyzed expression across different tissues including leaf, stem, flower, pollen, and seed to identify tissue-specific versus constitutive expression patterns [4].
• Replication: Proper biological replication (typically n=3) ensures statistical robustness in differential expression analysis, as demonstrated in banana transcriptome studies [83].
• Control Conditions: Parallel mock inoculations with sterile water provide baseline expression levels for distinguishing pathogen-induced responses from constitutive expression [83].

RNA-Seq Data Processing and Quality Control

Standardized processing pipelines ensure reproducible identification of constitutively expressed NBS genes:

• Quality Control: Tools like FastQC and MultiQC assess read quality, with typical thresholds of Q30 > 80% as used in banana blood disease research [83].
• Read Mapping: HISAT2 or similar aligners map reads to reference genomes, with alignment rates >70% typically considered acceptable [28].
• Transcript Quantification: Alignment-free tools like Salmon or alignment-dependent tools like Cufflinks calculate expression values (FPKM, TPM) [28] [83].
• Differential Expression: DESeq2 or Cuffdiff identify significantly differentially expressed genes using thresholds of log2FC > 1 and adjusted p-value ≤ 0.05 [83].

Table 3: Expression Analysis Tools and Applications for NBS Genes

Analysis Tool	Application	Key Features	NBS-Specific Applications
DESeq2 [83]	Differential expression	Negative binomial distribution, Wald test	Banana blood disease resistance [83]
Cufflinks/Cuffdiff [28]	Transcript assembly & differential expression	FPKM normalization, statistical testing	Nicotiana disease resistance studies [28]
qTeller [84]	Expression visualization	Gene model-specific expression data	Maize NBS gene expression analysis
Expression Atlas [85]	Multi-species expression data	Curated expression datasets	Cross-species comparisons

Defining Constitutive Expression Patterns

Constitutively expressed NBS genes demonstrate stable expression across multiple conditions:

• Stability Metrics: Genes with low coefficient of variation (<0.5) in FPKM/TPM values across conditions.
• Expression Thresholds: Minimum expression levels (FPKM >1) across majority of samples.
• Condition-Independence: Non-responsive to pathogen challenge, abiotic stresses, or developmental changes.

Research on cotton NBS genes identified orthogroups (OGs) with consistent expression patterns across susceptible and tolerant accessions under various biotic and abiotic stresses, suggesting constitutive roles in basal immunity [4].

Signaling Pathways and Experimental Workflows

NBS-Mediated Defense Signaling Pathways

NBS Genes in Plant Immunity

The diagram illustrates the central role of NBS-LRR genes in plant immune signaling pathways. Constitutively expressed NBS genes (highlighted in blue) function as key recognition receptors in effector-triggered immunity. CNL and TNL proteins directly or indirectly recognize pathogen effectors, while RNL proteins act as signal transducers downstream of multiple NLR receptors [82]. The integration of transcriptomic data enables identification of NBS genes maintaining stable expression across these defense pathways, suggesting fundamental roles in plant immunity.

Workflow for Identifying Constitutively Expressed NBS Genes

Constitutive NBS Gene Identification Pipeline

This workflow integrates genomic and transcriptomic data to filter constitutively expressed NBS genes. The process begins with genome-wide identification using HMMER and CDD searches, followed by RNA-seq data processing and quantification. The final filtering step applies thresholds for expression stability and magnitude across conditions to identify constitutively expressed NBS candidates [28] [83] [4].

Comparative Genomic Patterns of NBS Genes

Evolutionary Patterns Across Plant Families

NBS gene families exhibit diverse evolutionary patterns across plant species, influencing the identification of constitutively expressed members:

• Expansion Patterns: Rosaceae species show distinct evolutionary trajectories, with Rosa chinensis exhibiting "continuous expansion" while Fragaria vesca shows "expansion followed by contraction, then further expansion" [82].
• Lineage-Specific Differences: In Solanaceae, potato NBS-LRR genes show "consistent expansion," tomato displays "expansion followed by contraction," and pepper demonstrates a "shrinking" pattern [82].
• Allopolyploid Effects: Nicotiana tabacum, an allotetraploid, contains 603 NBS members—approximately the combined total of its parental species (N. sylvestris: 344, N. tomentosiformis: 279), with 76.62% traceable to parental genomes [28].

Orthogroup Analysis for Cross-Species Comparisons

Orthogroup (OG) analysis enables the identification of evolutionarily conserved NBS genes with potential constitutive expression:

• Core Orthogroups: Studies of 34 plant species identified 603 orthogroups, with OG0, OG1, and OG2 representing the most common across species [4].
• Expression Profiling: In cotton, OG2, OG6, and OG15 showed consistent upregulation across different tissues under various biotic and abiotic stresses in both susceptible and tolerant genotypes [4].
• Functional Validation: Virus-induced gene silencing (VIGS) of GaNBS (OG2) in resistant cotton demonstrated its role in virus tittering, confirming functional importance [4].

Table 4: NBS Gene Family Statistics Across Plant Species

Plant Species	Total NBS Genes	TNL Genes	CNL Genes	Other NBS	Study Year
Nicotiana tabacum	603	64 (TNL) + 9 (TN)	74 (CNL) + 150 (CN)	306 (NBS-only)	2025 [28]
Nicotiana sylvestris	344	37 (TNL) + 5 (TN)	48 (CNL) + 82 (CN)	172 (NBS-only)	2025 [28]
Nicotiana tomentosiformis	279	33 (TNL) + 7 (TN)	47 (CNL) + 65 (CN)	127 (NBS-only)	2025 [28]
Rosaceae (12 species)	2188	Variable	Variable	Variable	2022 [82]
34 plant species	12,820	Multiple classes	Multiple classes	168 domain architectures	2024 [4]

Bioinformatics Tools and Databases

Table 5: Essential Bioinformatics Resources for NBS Gene Analysis

Resource Category	Specific Resource	Application	Key Features
Genome Databases	NCBI Genome, Rosaceae.org, Banana Genome Hub	Genome assembly access	Annotated genomes, GFF files [28] [83] [82]
Domain Databases	PFAM, NCBI CDD	Domain identification	HMM profiles, conserved domains [28] [4]
Expression Databases	GEO, Expression Atlas, MaizeGDB	Transcriptomic data	RNA-seq datasets, visualization tools [86] [85] [84]
Analysis Tools	HMMER, OrthoFinder, MCScanX	Evolutionary analysis	Gene family identification, orthogrouping [28] [4]
Specialized Platforms	CottonFGD, MaizeGDB, IPF Database	Species-specific data	Curated expression datasets [4] [84]

• VIGS Vectors: Virus-Induced Gene Silencing systems for functional validation of candidate NBS genes, as demonstrated in cotton NBS studies [4].
• Pathogen Strains: Characterized isolates like Ralstonia syzygii subsp. celebesensis MY4101 for banana blood disease studies [83].
• RNA Extraction Kits: Commercial kits (e.g., RNeasy Plant Kit) for high-quality RNA isolation from plant tissues [83].
• qRT-PCR Reagents: Validation of RNA-seq results through quantitative real-time PCR with specific primers for target NBS genes [83].

The integration of transcriptomic data provides a powerful filtering approach for identifying constitutively expressed NBS genes that form the core components of plant immune systems across species. The comparative analysis presented here demonstrates that while NBS gene families exhibit remarkable diversity in size, architecture, and evolutionary patterns across plant lineages, computational pipelines combining HMMER searches, domain analysis, and RNA-seq profiling can effectively identify conserved, stably expressed family members. The resources, methodologies, and data frameworks outlined in this guide provide researchers with standardized approaches for cross-species comparison of NBS gene expression patterns, supporting ongoing efforts to understand the fundamental principles of plant immunity and accelerate the development of disease-resistant crop varieties through molecular breeding strategies.

Bridging Genomic Predictions with Functional Resistance Phenotypes

The nucleotide-binding site (NBS)-leucine-rich repeat (LRR) gene family constitutes one of the largest and most critical classes of plant resistance (R) genes, serving as fundamental components in plant innate immunity against diverse pathogens [4] [87]. These genes encode intracellular immune receptors that directly or indirectly recognize pathogen effectors, initiating robust defense signaling cascades culminating in effector-triggered immunity (ETI) [87] [20]. Expression profiling of NBS genes under pathogen challenge provides invaluable insights into the molecular basis of disease resistance, enabling the identification of key regulatory genes for crop improvement strategies [88] [89]. This comparative analysis synthesizes experimental data from multiple plant systems to delineate responsive NBS genes across pathogen interactions, presenting standardized methodologies for gene identification, expression analysis, and functional validation. By integrating findings from recent transcriptomic studies, we aim to establish a cross-species framework for understanding NBS gene regulation during plant defense responses, providing researchers with validated experimental approaches and analytical tools for investigating this crucial gene family.

NBS Gene Family: Structural Diversity and Classification

The NBS-LRR gene family represents the most prevalent class of plant R genes, characterized by a conserved nucleotide-binding adaptor shared by APAF-1, R proteins, and CED-4 (NB-ARC) domain and C-terminal leucine-rich repeats [4] [68]. Based on N-terminal domain architecture, NBS-encoding genes are primarily classified into three major subfamilies: TIR-NBS-LRR (TNL) containing Toll/interleukin-1 receptor domains, CC-NBS-LRR (CNL) featuring coiled-coil domains, and RPW8-NBS-LRR (RNL) with resistance to powdery mildew8 domains [68] [20]. The structural organization of these domains dictates their functional specialization, with TNL and CNL proteins primarily responsible for pathogen recognition, while RNL proteins facilitate downstream defense signal transduction [68].

Genome-wide comparative analyses reveal remarkable diversity in NBS gene composition across plant species. A comprehensive study examining 34 plant species identified 12,820 NBS-domain-containing genes, classifying them into 168 distinct classes based on domain architecture patterns [4]. These encompass both classical configurations (NBS, NBS-LRR, TIR-NBS, TIR-NBS-LRR) and species-specific structural combinations (TIR-NBS-TIR-Cupin1-Cupin1, TIR-NBS-Prenyltransf, Sugar_tr-NBS) [4]. The number of NBS genes exhibits substantial interspecies variation, ranging from 73 identified in Akebia trifoliata to over 2,000 in some flowering plants [68]. This expansion primarily results from tandem and whole-genome duplication events, with Brassica species exhibiting species-specific gene amplification through tandem duplication following divergence from Arabidopsis thaliana [14].

Table 1: NBS-LRR Gene Family Composition Across Plant Species

Plant Species	Total NBS Genes	CNL	TNL	RNL	Reference
Akebia trifoliata	73	50	19	4	[68]
Arabidopsis thaliana	167	51	-	-	[4] [14]
Brassica oleracea	157	-	-	-	[14]
Brassica rapa	206	-	-	-	[14]
Passiflora edulis (purple)	25	25	0	0	[20]
Passiflora edulis (yellow)	21	21	0	0	[20]

Chromosomal distribution patterns consistently show NBS genes frequently clustered at chromosome termini, with both homogeneous and heterogeneous arrangements [68] [14]. For instance, in A. trifoliata, 64 mapped NBS candidates distributed unevenly across 14 chromosomes, with 41 genes located in clusters and 23 as singletons [68]. Evolutionary analyses indicate tandem and dispersed duplications as primary mechanisms for NBS gene expansion, producing 33 and 29 genes respectively in A. trifoliata [68]. The evolutionary trajectory of NBS genes following whole-genome triplication in Brassica ancestors reveals rapid deletion or loss of triplicated homologous gene pairs, followed by lineage-specific tandem duplication [14].

Experimental Designs for Expression Profiling

Comparative Transcriptomics of Resistant and Susceptible Genotypes

A powerful approach for identifying pathogen-responsive NBS genes involves comparative transcriptomic analysis of genotypes with contrasting resistance phenotypes under pathogen challenge. This design enables researchers to distinguish defense-associated expression patterns from general stress responses. In peanut (Arachis hypogaea) infected with Agroathelia rolfsii, RNA sequencing of resistant (Georgia-03L) and susceptible (Valencia C) genotypes identified strong induction of NBS-LRR resistance genes along with receptor-like kinases and transcription factors in the resistant line [89]. Similarly, grapevine transcriptome analysis of cultivars with differential susceptibility to grapevine trunk diseases (GTDs) revealed 64 differentially expressed genes (DEGs) associated with symptomatology regardless of cultivar [88].

The experimental workflow typically involves controlled pathogen inoculation, tissue sampling at strategic time points, RNA extraction and quality control, library preparation and sequencing, followed by bioinformatic analysis. For peanut stem rot resistance studies, researchers inoculated 52-day-old plants with A. rolfsii mycelial slurry, collecting stem samples at 72 hours post-inoculation (hpi) from the lower portion of the main stem [89]. Rigorous RNA quality control measures are implemented, accepting only samples with RNA Integrity Number (RIN) ≥ 8.0 for subsequent library preparation and sequencing [89].

Time-Series Expression Analysis

Temporal monitoring of NBS gene expression provides insights into the dynamics of defense activation and the hierarchical organization of immune signaling. Transcriptome profiling of starry flounder (Platichthys stellatus) following Streptococcus parauberis infection demonstrated a temporal shift in immune response, with early activation of DNA damage repair pathways (3 hpi) transitioning to immune modulation and energy conservation (48 hpi) [90]. Although this example comes from animal immunity, similar temporal dynamics occur in plant systems, where early transcriptional responses often involve pathogen recognition receptors and signaling components, while later responses may involve amplification of defense signals and systemic immunity.

In passion fruit, transcriptome data indicated that PeCNL3, PeCNL13, and PeCNL14 were differentially expressed under Cucumber mosaic virus infection and cold stress, suggesting these genes may function in multiple stress response pathways [20]. Time-series expression data are particularly valuable for distinguishing primary response genes from secondary responders in defense networks, potentially identifying key regulatory nodes within NBS signaling networks.

Tissue-Specific Expression Profiling

Spatial expression patterns of NBS genes provide critical information about their site of action and potential functional specialization. In A. trifoliata, transcriptome analysis of three fruit tissues (rind, flesh, and seed) across four developmental stages revealed that NBS genes were generally expressed at low levels, with a subset showing relatively high expression during later development in rind tissues [68]. This tissue-specific expression pattern suggests specialized defensive roles in particular organs or developmental stages.

Comparative analysis of immune responses across tissues in starry flounder demonstrated that liver tissue exhibited greater transcriptional variability following infection, indicating its role in systemic immune regulation, while leukocytes primarily contributed to pathogen recognition [90]. In plant systems, similar compartmentalization of defense functions occurs, with some NBS genes showing root-specific expression while others are leaf-predominant, reflecting adaptation to tissue-specific pathogen challenges.

Key Methodologies and Protocols

Identification and Classification of NBS Genes

Standardized protocols for NBS gene identification employ a combination of homology searches and domain verification. The typical workflow begins with BLASTP analysis using reference NBS protein sequences (e.g., NB-ARC domain PF00931) against target proteomes [68] [20]. Candidate sequences are subsequently verified using hidden Markov model (HMM) profiling with tools like HMMER, applying trusted cutoff thresholds [4] [14]. For example, in the identification of 12,820 NBS genes across 34 species, researchers used PfamScan.pl HMM search script with default e-value (1.1e-50) and background Pfam-A_hmm model [4].

Domain architecture analysis forms the basis for NBS gene classification. The presence of TIR (PF01582), RPW8 (PF05659), and LRR (PF08191) domains is typically determined using the NCBI Conserved Domain Database, while coiled-coil domains are identified using tools like Paircoil2 or MARCOIL with appropriate probability thresholds [68] [14]. Classification systems organize genes into classes based on similar domain architectures, enabling comparative analysis across species [4].

Figure 1: NBS Gene Identification and Classification Workflow

Transcriptome Sequencing and Analysis

RNA sequencing represents the current gold standard for comprehensive expression profiling. Experimental protocols typically involve RNA extraction from pathogen-challenged tissues, quality assessment, library preparation, and high-throughput sequencing. In peanut studies, total RNA was extracted using commercial kits (e.g., Spectrum Plant Total RNA Kit) with on-column DNase I treatment to remove genomic DNA contamination [89]. Quality-controlled RNA (RIN > 8.0) was used to construct poly-A-enriched libraries sequenced on platforms such as DNBSEQ-T7 or Illumina systems [89].

Bioinformatic processing includes quality filtering, read alignment, differential expression analysis, and functional annotation. For peanut transcriptomics, researchers filtered raw data using SOAPnuke to remove adapter sequences and low-quality reads, then aligned clean reads to reference genomes using HISAT2 [89]. Differential expression analysis employing tools like DESeq2 or edgeR identifies significantly regulated genes under pathogen challenge, with subsequent functional annotation through databases such as GO, KEGG, and Pfam [88] [89].

Functional Validation Approaches

Functional validation of candidate NBS genes typically employs genetic approaches to establish their role in disease resistance. Virus-induced gene silencing (VIGS) provides an efficient method for transient gene knockdown to assess gene function. In cotton, silencing of GaNBS (OG2) in resistant plants through VIGS demonstrated its putative role in virus tittering, establishing its importance in resistance to cotton leaf curl disease [4].

Heterologous expression in model systems and stable transformation of susceptible genotypes offer complementary validation strategies. While not explicitly detailed in the surveyed studies, these approaches are widely used in the field to confirm the function of putative NBS resistance genes. Additionally, protein interaction studies such as yeast two-hybrid screening and bimolecular fluorescence complementation can elucidate signaling mechanisms, as demonstrated by interactions between NBS proteins and pathogen effectors [87].

Expression Profiles of Responsive NBS Genes

Orthogroup Expression Patterns

Comparative analysis of NBS gene expression across species reveals conserved orthogroups with pathogen-responsive profiles. A comprehensive study examining NBS genes across 34 plant species identified 603 orthogroups (OGs), including core orthogroups (OG0, OG1, OG2) common across multiple species and unique orthogroups (OG80, OG82) specific to particular lineages [4]. Expression profiling demonstrated putative upregulation of OG2, OG6, and OG15 in different tissues under various biotic and abiotic stresses in susceptible and tolerant cotton genotypes responding to cotton leaf curl disease (CLCuD) [4].

Table 2: Expression Profiles of NBS Genes Under Pathogen Challenge

Plant System	Pathogen	Responsive NBS Genes	Expression Pattern	Reference
Cotton	Cotton leaf curl virus	OG2, OG6, OG15	Upregulated in tolerant genotypes	[4]
Peanut	Agroathelia rolfsii	NBS-LRR genes	Strongly induced in resistant genotype	[89]
Passion fruit	Cucumber mosaic virus	PeCNL3, PeCNL13, PeCNL14	Differentially expressed	[20]
Grapevine	Grapevine trunk diseases	Multiple NBS genes	Varied by cultivar susceptibility	[88]
Akebia trifoliata	Developmental regulation	Subset of NBS genes	Higher in rind during late development	[68]

The genetic architecture of resistance often involves specific NBS gene variants. Comparison between susceptible (Coker 312) and tolerant (Mac7) Gossypium hirsutum accessions identified numerous unique variants in NBS genes, with Mac7 exhibiting 6,583 variants compared to 5,173 in Coker312 [4]. These sequence variations potentially affect protein function and pathogen recognition specificity, contributing to contrasting resistance phenotypes.

Co-expression Network Analysis

Weighted Gene Co-expression Network Analysis (WGCNA) identifies modules of coordinately expressed genes associated with resistance traits. In peanut resistance to A. rolfsii, WGCNA identified a co-expression module enriched with genes involved in oxidative stress response, secondary metabolism, and cell wall reinforcement [89]. Although not exclusively containing NBS genes, such defense-related modules often include NBS genes as key nodes, potentially representing coordinated immune signaling networks.

Integration of expression data with genomic localization can reveal regulatory mechanisms. For instance, cis-element analysis of passion fruit CNL genes identified elements involved in plant growth, hormones, and stress response, providing insights into potential regulatory mechanisms governing their expression patterns [20]. Such integrated analyses help establish connections between genetic sequences, regulatory elements, and expression dynamics in plant immunity.

Signaling Pathways and Molecular Interactions

NBS-LRR proteins function as central components in plant immune signaling networks, detecting pathogen effectors through direct or indirect recognition mechanisms [87]. Direct effector binding provides the most straightforward recognition mechanism, exemplified by interactions between rice Pi-ta protein and fungal effector AVR-Pita, flax L proteins and fungal AvrL567 effectors, and Arabidopsis RRS1 and bacterial PopP2 [87]. Indirect recognition occurs through guard mechanisms, where NBS-LRR proteins monitor the status of host proteins targeted by pathogen effectors, as demonstrated by Arabidopsis RPM1 and RPS2 surveillance of RIN4 protein modifications [87].

Figure 2: NBS-LRR Activation Mechanisms in Plant Immunity

Upon pathogen recognition, NBS-LRR proteins undergo conformational changes facilitating ADP-to-ATP exchange, transitioning to activated states that initiate downstream signaling [87]. Structural studies indicate that LRR domains form solenoid-like structures with parallel β-sheets lining inner concave surfaces, potentially mediating protein-protein interactions critical for effector recognition and signal transduction [87]. Activation of NBS-LRR proteins triggers defense signaling networks including MAPK cascades, calcium signaling, reactive oxygen species production, and hormonal pathways, collectively establishing antimicrobial environments and enhancing resistance to subsequent infections [88] [89].

Protein interaction studies provide mechanistic insights into NBS function. Molecular docking analyses demonstrate strong interactions between putative NBS proteins and ADP/ATP molecules, reflecting their nucleotide-binding capacity, as well as with core proteins of the cotton leaf curl disease virus, suggesting potential recognition mechanisms [4]. Such molecular interactions underlie the immune activation process that ultimately restricts pathogen proliferation.

Research Reagent Solutions

Table 3: Essential Research Reagents for NBS Gene Expression Studies

Reagent Category	Specific Products/Tools	Application	Reference
RNA Extraction Kits	Spectrum Plant Total RNA Kit	High-quality RNA isolation from plant tissues	[89]
Library Prep Kits	Poly-A enrichment kits	mRNA sequencing library construction	[89]
Sequencing Platforms	DNBSEQ-T7, Illumina	High-throughput transcriptome sequencing	[88] [89]
Alignment Tools	HISAT2, SOAPnuke	Read alignment and quality processing	[89]
Domain Databases	Pfam, CDD, InterPro	NBS domain identification and verification	[68] [20]
Expression Analysis	DESeq2, edgeR	Differential expression analysis	[88]
Co-expression Analysis	WGCNA	Identification of correlated gene modules	[89]
Functional Annotation	GO, KEGG, PlantCyc	Pathway enrichment and functional classification	[88] [89]

Additional specialized reagents include commercial growing media like Metro-Mix 840 for standardized plant growth [89], acidified potato dextrose agar for fungal pathogen culture [89], and specific computational tools for phylogenetic analysis (OrthoFinder, FastTree) and motif identification (MEME Suite) [4] [68]. Standardized pathogen inoculation materials, such as fungal mycelial slurries for soil-borne pathogens [89] or viral inocula for leaf infections [4], ensure consistent challenge conditions across experiments. For functional validation, VIGS vectors provide efficient tools for transient gene silencing in numerous plant species [4].

Expression profiling of NBS genes under pathogen challenge has illuminated the dynamic regulation of this crucial gene family in plant immunity. Comparative analyses across diverse pathosystems reveal both conserved and species-specific expression patterns, highlighting the evolutionary innovation in plant immune systems. The identification of responsive NBS genes, particularly those consistently upregulated across multiple resistance interactions, provides valuable candidates for crop improvement programs.

The experimental approaches and methodologies reviewed here offer standardized frameworks for investigating NBS gene regulation, from comprehensive identification and classification to functional validation. Integration of transcriptomic data with genomic, genetic, and protein interaction analyses provides multidimensional insights into NBS gene function. These research strategies have already yielded practical applications, including the development of molecular markers for resistance breeding and the identification of candidate genes for genetic engineering. As genomic technologies continue advancing, expression profiling of NBS genes will undoubtedly uncover additional layers of complexity in plant immune networks, further enabling the development of durable disease resistance in agricultural systems.

Functional validation is a critical step in plant genomics, bridging the gap between gene prediction and demonstrated biological function. For nucleotide-binding site-leucine-rich repeat (NBS-LRR) genes—the largest class of plant disease resistance (R) genes—several powerful approaches have been developed to confirm gene function and elucidate mechanisms of pathogen recognition and immune signaling [4] [24]. This guide provides a comparative analysis of three central methodologies: virus-induced gene silencing (VIGS), heterologous expression, and mutagenesis. Within the expanding field of comparative genomics, where thousands of NBS-encoding genes have been identified across species [4] [9] [91], selecting the appropriate validation strategy is paramount for accurately characterizing the role of these genes in plant immunity.

Comparative Analysis of Functional Validation Methods

The table below summarizes the key characteristics, applications, and outputs of the three primary functional validation approaches used in plant NBS-LRR gene research.

Table 1: Comparison of Major Functional Validation Approaches for Plant NBS-LRR Genes

Feature	VIGS (Virus-Induced Gene Silencing)	Heterologous Expression	Mutagenesis
Core Principle	Post-transcriptional gene silencing using recombinant viral vectors [92]	Expressing a target gene in a different, susceptible host species [91]	Disrupting target gene function via chemical or genome editing tools [93]
Primary Application	Rapid loss-of-function analysis to assess gene necessity [4] [94]	Gain-of-function analysis to test gene sufficiency for resistance [91]	Confirming gene identity and studying structure-function relationships [93]
Typical Workflow Duration	3-8 weeks post-inoculation [92]	Several months (including transformation) [91]	3-6 months for screening (e.g., EMS) [93]
Key Readouts	Phenotypic susceptibility, pathogen titers, downregulation of target transcript [4] [94]	Hypersensitive response (HR), pathogen growth restriction [91]	Loss-of-resistance phenotype, identification of premature stop codons/missense mutations [93]
Throughput	Medium to High [92]	Low to Medium [91]	High (for EMS populations) [93]
Technical Complexity	Moderate (requires vector engineering and plant inoculation) [92]	High (requires stable transformation) [92]	Low (EMS) to High (CRISPR/Cas9) [93]

Detailed Experimental Protocols

Protocol 1: Virus-Induced Gene Silencing (VIGS)

VIGS is a powerful reverse-genetics tool that leverages the plant's RNAi machinery to knock down endogenous gene expression. The following protocol is adapted from studies in cotton and pepper [4] [92] [94].

• Insert Selection and Vector Construction: A unique, 250-400 base pair fragment of the target gene (e.g., an NBS-LRR like GaNBS or CaAN2) is amplified from cDNA [4] [94]. This fragment is cloned into a VIGS vector, most commonly the Tobacco Rattle Virus (TRV)-based pTRV2 vector.
• Transformation and Agroinfiltration: The recombinant pTRV2 vector and a helper vector (pTRV1) are introduced into Agrobacterium tumefaciens. The bacterial cultures are grown, resuspended in an induction medium (e.g., with acetosyringone), and infiltrated into the leaves of young plants, typically at the 2-4 leaf stage [92] [94].
• Phenotypic Analysis: After 3-4 weeks, silencing efficacy is assessed. For genes involved in visible processes (e.g., CaPDS in pigment biosynthesis), photobleaching provides a visual marker [94]. For R genes, silenced plants are challenged with a pathogen, and disease susceptibility is scored.
• Molecular Validation: Silencing is confirmed using quantitative RT-PCR (qRT-PCR) to measure the reduction in target gene mRNA levels. Pathogen biomass in control versus silenced plants can be quantified to confirm the role of the targeted gene in resistance [4] [94].

VIGS Experimental Workflow

Protocol 2: Heterologous Expression

This approach tests whether a candidate R gene is sufficient to confer resistance in a susceptible plant background [91].

• Gene Cloning and Vector Construction: The full-length coding sequence (CDS) of the candidate NBS-LRR gene is amplified and cloned into a stable expression vector under a strong constitutive promoter (e.g., CaMV 35S) [91].
• Plant Transformation and Selection: The construct is introduced into a susceptible plant model (e.g., Nicotiana benthamiana or Arabidopsis thaliana) using Agrobacterium-mediated transformation. Transgenic lines are selected using antibiotics or herbicides, and homozygotic T2 or T3 generations are established.
• Resistance Phenotyping: Transgenic and control plants are inoculated with the relevant pathogen. Resistance is evaluated by monitoring for the development of a hypersensitive response (HR), a rapid, localized cell death at the infection site, and/or by measuring reduced pathogen growth compared to control plants [91].
• Expression Confirmation: The expression of the transgene in the resistant transgenic lines is confirmed via RT-PCR or Western blotting.

Protocol 3: Mutagenesis

Mutagenesis creates genetic alterations to disrupt gene function. Both chemical and targeted methods are widely used [93].

• Population Generation:
  • EMS Mutagenesis: Seeds are treated with ethyl methanesulfonate (EMS), which induces random G/C to A/T point mutations throughout the genome. Treated seeds (M0) are grown, and the subsequent M2 generation is used for forward genetic screens [93].
  • CRISPR/Cas9: Single-guide RNAs (sgRNAs) are designed to target specific exons of the candidate gene. A CRISPR/Cas9 construct is assembled and used to transform plants [93].
• Mutant Screening:
  • Forward Screen (EMS): M2 plants are screened for a loss-of-resistance phenotype (e.g., susceptibility to a pathogen). This is efficient in wheat, as the polyploid genome can tolerate high mutation rates, and most loss-of-function mutants map directly to the R gene rather than redundant signaling components [93].
  • Reverse Screen (CRISPR): Transformed plants are genotyped to identify individuals with insertions or deletions (indels) in the target gene.
• Gene Identification and Validation:
  • For EMS mutants, bulk segregant analysis and whole-genome sequencing (e.g., MutMap) or RNA-Seq of pooled mutants (MutIsoSeq) is used to pinpoint the causal mutation [93]. Sanger sequencing confirms the mutation in individual mutants.
  • For CRISPR mutants, the susceptibility of the gene-edited lines is confirmed through pathogen assays.

Key Signaling Pathways and Genetic Relationships

NBS-LRR genes are central components of Effector-Triggered Immunity (ETI). The diagram below illustrates the simplified signaling logic of how these genes are validated functionally.

The Scientist's Toolkit: Essential Research Reagents

The table below lists critical reagents and materials required for the functional validation experiments described in this guide.

Table 2: Key Research Reagents for Functional Validation of NBS-LRR Genes

Reagent/Material	Function/Application	Example Use Cases
TRV VIGS Vectors (pTRV1, pTRV2)	RNA virus-based system for inducing gene silencing; bipartite system for broad-host-range application [92]	Silencing CaPDS in pepper as a visual marker; validating role of GaNBS in cotton virus resistance [4] [94]
Agrobacterium tumefaciens (e.g., GV3101)	Delivery vehicle for introducing DNA constructs (VIGS vectors, heterologous expression, CRISPR) into plant cells [92] [94]	Agroinfiltration for transient VIGS; stable transformation for heterologous expression
Ethyl Methanesulfonate (EMS)	Chemical mutagen that induces random point mutations (G/C to A/T) for forward genetics screens [93]	Generating large mutant populations in wheat to identify loss-of-function mutants for R genes like Sr6 [93]
CRISPR/Cas9 System	Genome editing tool for targeted gene knock-out via double-strand breaks and error-prone repair [93]	Creating precise knock-out mutants of the Sr6 gene in wheat to confirm its function [93]
Phytohormones & Selection Agents	Antibiotics for bacterial and plant selection; plant hormones for regeneration (e.g., in transformation) [92]	Selecting transformed plants during heterologous expression and genome editing

Plant diseases pose a significant threat to global crop yield and quality. Understanding the genetic basis of disease resistance is paramount for developing resilient crop varieties. Nucleotide-binding site (NBS) domain genes constitute one of the largest families of plant resistance (R) genes, playing a critical role in effector-triggered immunity (ETI) by recognizing diverse pathogen effectors [95] [4]. This guide employs a comparative genomics approach to objectively analyze the architecture, evolution, and functional mechanisms of NBS-encoding genes in two industrially significant plants: tung tree (Vernicia fordii) and cotton (Gossypium spp.). By dissecting the genetic differences between susceptible and resistant varieties, we provide a framework for understanding disease resistance mechanisms and inform future breeding strategies.

Genome-Wide Analysis of NBS-Encoding Genes

Identification and Classification

Comprehensive genome-wide analyses have revealed significant differences in the number and type of NBS-encoding genes between susceptible and resistant varieties of cotton and tung tree.

Table 1: NBS-Encoding Gene Profiles in Cotton and Tung Tree

Species/Variety	Total NBS Genes	CNL	TNL	Other NBS Types	Key Characteristics
G. raimondii (Resistant diploid)	365 [12]	29.32% [12]	Higher proportion [12]	RNL: ~2% [12]	High proportion of TNL genes [12]
G. barbadense (Resistant tetraploid)	682 [12]	Lower proportion than susceptible [12]	Higher proportion [12]	RNL: ~2% [12]	Inherits more NBS genes from G. raimondii [12]
G. arboreum (Susceptible diploid)	246 [12]	32.52% [12]	Lower proportion [12]	RNL: ~2% [12]	Higher proportion of CN and N genes [12]
G. hirsutum (Susceptible tetraploid)	588 [12]	Higher proportion than resistant [12]	Lower proportion [12]	RNL: ~2% [12]	Inherits more NBS genes from G. arboreum [12]
Vernicia fordii (Tung Tree)	1 candidate identified [96]	Specific type not detailed	Specific type not detailed	Involved in flavonoid biosynthesis [96]	NBS-LRR candidate gene for Fusarium resistance [96]

In cotton, the allotetraploid species (G. hirsutum and G. barbadense) possess nearly double the number of NBS genes compared to their diploid progenitors, a consequence of hybridization and subsequent gene duplication or loss [12]. A key finding is the asymmetric evolution of NBS-encoding genes. The resistant tetraploid G. barbadense inherited a larger proportion of its NBS genes from the resistant D-genome progenitor G. raimondii, whereas the susceptible tetraploid G. hirsutum inherited more from the susceptible A-genome progenitor G. arboreum [12]. This inheritance pattern is particularly evident in the distribution of TIR-NBS-LRR (TNL) genes, which are about seven times more abundant in the resistant G. raimondii and G. barbadense compared to their susceptible counterparts [12].

Structural and Evolutionary Diversification

NBS-encoding genes exhibit considerable structural diversity. They can be classified into "regular" genes, which contain all five conserved NBS motifs (P-loop, kinase-2, kinase-3a, GLPL, and MHDL), and "non-regular" genes, which possess only some of these motifs [95]. A prominent feature of NBS gene evolution is their tendency to form clusters on chromosomes, often resulting from tandem and segmental duplications [4] [12] [97]. For instance, in a resistant cultivar of G. barbadense, 37.5% of identified CC-NBS-LRR (CNL) genes were organized into 12 gene clusters [97]. These clusters act as genetic variation libraries, fostering the evolution of new resistance specificities through recombination and diversifying selection [95] [97].

Figure 1: NBS Gene Classification and Domain Architecture. NBS-encoding resistance genes are primarily classified into TNL, CNL, and RNL types based on their N-terminal domains (TIR, CC, or RPW8). All types share a central NBS domain for nucleotide binding and a C-terminal LRR domain for pathogen recognition.

Experimental Methodologies for Functional Validation

Genome-Wide Identification and Bioinformatics Analysis

Protocol 1: Identification and Classification of NBS-Encoding Genes

• Data Retrieval: Obtain the latest genome assemblies and protein sequence files for the target species from databases such as NCBI, Phytozome, or Plaza [4] [1].
• HMMER Search: Use HMMER software (e.g., HMMER3) with a hidden Markov model (HMM) profile of the NB-ARC domain (PF00931) to scan the proteome for candidate genes [12] [1]. A stringent e-value cut-off (e.g., 1.1e-50) is recommended [4].
• Domain Validation: Subject candidate sequences to domain analysis tools like InterProScan, PfamScan, SMART, and MARCOIL to confirm the presence of the NBS domain and identify associated domains (TIR, CC, LRR) [95] [1].
• Classification and Analysis: Classify genes based on domain architecture (e.g., CNL, TNL, NL). Subsequently, perform phylogenetic analysis, motif discovery, chromosomal location mapping, and synteny analysis to understand evolutionary relationships and genomic distribution [4] [12].

Association Studies and Candidate Gene Discovery

Protocol 2: Genome-Wide Association Study (GWAS) for Disease Resistance

• Phenotyping: Evaluate a natural population or association panel for disease resistance in multiple environments (e.g., greenhouse and field) with several replicates. The disease index (DI) is commonly used to quantify symptoms [98] [99].
• Genotyping: Utilize high-throughput sequencing technologies like Specific-locus Amplified Fragment Sequencing (SLAF-seq) or Genotyping-by-Sequencing (GBS) to generate thousands of single nucleotide polymorphisms (SNPs) across the panel [98].
• Association Analysis: Perform trait-SNP association analysis using mixed linear models to correct for population structure. Significance thresholds are often set based on a Bonferroni correction (e.g., ( P < 1/n ), where ( n ) is the number of SNPs) [98].
• Candidate Gene Identification: Based on significant SNP loci, define haplotype blocks and identify genes within or near these associated genomic regions. Prioritize candidates that encode known resistance protein domains (e.g., TIR-NBS-LRR) [98].

Functional Characterization Using Virus-Induced Gene Silencing (VIGS)

Protocol 3: Functional Validation via VIGS

• Vector Construction: Clone a 200-300 bp fragment of the candidate gene into a VIGS vector (e.g., derived from Tobacco Rattle Virus, pTRV2) [98] [4] [97].
• Plant Infiltration: Mix the recombinant pTRV2 vector with the helper strain (pTRV1) and infiltrate into cotyledons or true leaves of young plants (e.g., cotton) using Agrobacterium-mediated transformation [98] [97].
• Phenotypic Validation: After successful gene silencing (confirmed by qRT-PCR), challenge the plants with the pathogen (e.g., Verticillium dahliae). Compare the disease symptoms in silenced plants to control plants (e.g., infiltrated with empty vector) [98] [97].
• Defense Response Analysis: Measure defense-related parameters in silenced and control plants, such as the accumulation of reactive oxygen species (ROS), the expression of pathogenesis-related (PR) genes, and the levels of defense hormones like salicylic acid (SA) [97].

Signaling Pathways and Defense Mechanisms

The defense responses mediated by NBS-encoding genes are complex and involve specific signaling pathways. In cotton, the CNL protein GbCNL130 confers resistance to Verticillium wilt by activating the salicylic acid (SA)-dependent pathway. This leads to a strong oxidative burst and upregulation of PR genes, creating a hostile environment for the pathogen [97]. In contrast, research in tung tree has highlighted a distinct resistance mechanism centered on flavonoid biosynthesis. The UDP-glycosyltransferase VfUGT90A2, a key hub gene induced upon Fusarium infection, glycosylates flavonoid compounds like quercetin. This process enhances the production of antifungal metabolites such as quercitrin and myricitrin, which directly inhibit pathogen growth [96].

Figure 2: Comparative Defense Signaling Pathways. Resistant cotton varieties often employ CNL proteins to activate SA-dependent defense signaling, leading to ROS and PR gene expression. Tung tree resistance can involve UGT-mediated flavonoid glycosylation to produce direct antifungal compounds.

The Scientist's Toolkit: Essential Research Reagents and Solutions

Table 2: Key Reagents and Solutions for Comparative Genomics of Plant Disease Resistance

Reagent/Solution	Function/Application	Example Use Case
HMMER Suite	Identifies protein domains (e.g., NB-ARC PF00931) using hidden Markov models.	Genome-wide identification of NBS-encoding genes [12].
InterProScan/Pfam	Scans protein sequences against multiple domain databases for functional annotation.	Validating NBS domain presence and classifying R genes into CNL/TNL [95] [1].
TRV-based VIGS Vectors (pTRV1, pTRV2)	Virus-Induced Gene Silencing system for rapid loss-of-function studies in plants.	Functional validation of candidate R genes like GaNBS and GbCNL130 [98] [4] [97].
GWAS Analysis Pipelines	Statistically associates genomic markers (SNPs) with phenotypic traits.	Mapping Verticillium wilt resistance loci in natural cotton populations [98] [99].
ClustalW/MEGA	Performs multiple sequence alignment and phylogenetic tree construction.	Evolutionary analysis and orthogrouping of NBS genes across species [95] [4].

This comparative guide elucidates the genomic foundations of disease resistance in tung tree and cotton. The evidence demonstrates that resistant varieties are characterized by distinct NBS-encoding gene profiles, particularly a enrichment of TNL-type genes in cotton, and the deployment of both NBS and non-NBS resistance mechanisms, such as flavonoid glycosylation in tung tree. The asymmetric evolution of NBS genes in allopolyploid cotton, where the resistant tetraploid G. barbadense preferentially retained NBS genes from its resistant D-genome progenitor, provides a powerful explanation for observed interspecific differences in disease susceptibility. The experimental protocols and reagents detailed herein provide a roadmap for researchers to further dissect these complex traits. Future research leveraging these comparative genomics insights will accelerate the development of disease-resistant crop varieties through marker-assisted selection and genetic engineering.

Plant immunity relies on a sophisticated surveillance system where intracellular nucleotide-binding leucine-rich repeat receptors (NLRs) play a critical role in detecting pathogen effectors and initiating robust defense responses [100]. These proteins typically contain a conserved nucleotide-binding site (NBS) domain and a C-terminal leucine-rich repeat (LRR) region, which facilitate pathogen recognition and immune signaling activation [9]. Based on their N-terminal domains, NLRs are classified into distinct subfamilies: CNLs (containing coiled-coil domains), TNLs (with Toll/interleukin-1 receptor domains), and RNLs (featuring RPW8 domains) [100] [9].

The domestication of crop species has frequently selected for traits favoring yield and quality, sometimes at the expense of natural defense mechanisms. Garden asparagus (Asparagus officinalis), recognized as the "king of vegetables" in international markets, provides an excellent system for investigating how artificial selection has shaped NLR gene evolution [100] [9]. This guide presents a comparative analysis of NLR gene repertoires between cultivated asparagus and its wild relatives, integrating quantitative genomic data, experimental methodologies, and functional insights to elucidate the genetic consequences of domestication on plant immunity.

Comparative Genomic Analysis Reveals NLR Contraction During Domestication

Comprehensive genome-wide identification of NLR genes across Asparagus species reveals a striking pattern of gene family contraction associated with domestication. Wild relatives maintain substantially larger and more diverse NLR repertoires compared to the cultivated species [100] [9].

Table 1: NLR Gene Distribution in Asparagus Species

Species	Domestication Status	Total NLR Genes	CNL Subfamily	TNL Subfamily	RNL Subfamily	Other/Truncated
A. setaceus	Wild	63	35	18	2	8
A. kiusianus	Wild	47	29	12	1	5
A. officinalis	Cultivated	27	19	5	1	2

Table 2: Orthologous NLR Gene Conservation Between A. setaceus and A. officinalis

Conservation Category	Gene Count	Percentage	Functional Status in A. officinalis
Conserved orthologous pairs	16	25.4%	Reduced or unresponsive expression
NLRs lost in domestication	47	74.6%	Complete gene loss
Retained NLRs with downregulation	12	75%	Impaired defense signaling
Retained NLRs with unchanged expression	3	18.8%	Non-responsive to pathogen challenge
Retained NLRs with upregulated expression	1	6.2%	Potentially functional

The genomic data reveal a clear trend: cultivated asparagus has experienced a 57% reduction in NLR genes compared to A. setaceus and a 42% reduction compared to A. kiusianus [100]. This contraction affects all NLR subfamilies but appears most pronounced in the TNL class, potentially narrowing the spectrum of pathogen recognition capabilities in the domesticated species [100] [9].

Orthologous analysis identified only 16 conserved NLR gene pairs between A. setaceus and A. officinalis, representing the core NLR repertoire preserved during domestication [100]. The massive loss of NLR diversity (approximately 75% of wild NLRs) likely contributes to the enhanced disease susceptibility observed in cultivated asparagus, particularly toward fungal pathogens like Phomopsis asparagi [100] [9].

Methodological Framework for NLR Gene Identification and Characterization

Genome-Wide NLR Identification and Classification

The comparative analysis of NLR genes across Asparagus species employed a rigorous computational pipeline to ensure comprehensive identification and accurate classification [100]:

• HMMER Searches: Initial identification used Hidden Markov Model (HMM) searches with the conserved NB-ARC domain (Pfam: PF00931) as query, applying an E-value cutoff of 1e-10 [100].
• BLAST Validation: Complementary BLASTp analyses against reference NLR proteins from Arabidopsis thaliana, Oryza sativa, and Allium sativum provided validation through sequence similarity [100].
• Domain Architecture Verification: Candidate sequences underwent thorough domain characterization using InterProScan and NCBI's Batch CD-Search, retaining only sequences containing the NB-ARC domain (E-value ≤ 1e-5) as bona fide NLR genes [100].
• Final Classification: Genes were categorized into subfamilies (CNL, TNL, RNL, and truncated variants) based on their complete domain architecture using the Pfam and PRGdb 4.0 databases [100].

Phylogenetic and Evolutionary Analysis

Reconstructing evolutionary relationships among NLR genes employed these methodological approaches:

• Multiple Sequence Alignment: Protein sequences of candidate NLR genes were consolidated and aligned using Clustal Omega [100].
• Phylogenetic Tree Construction: Maximum likelihood trees were built using MEGA software based on the JTT matrix-based model, with bootstrap testing of 1000 replicates to assess node support [100].
• Orthogroup Analysis: Orthologous genes between species were clustered using OrthoFinder v2.2.7, which normalized BLAST bit scores based on gene length and phylogenetic distance [100].
• Collinearity Analysis: "One Step MCScanX" from TBtools enabled detection of syntenic blocks and comparative genomic architecture across species [100].

Expression Profiling and Functional Validation

The functional assessment of NLR genes utilized both computational and experimental approaches:

• Cis-Element Analysis: The PlantCARE database identified defense-related and phytohormone-responsive elements in promoter regions (2000 bp upstream of start codons) [100].
• Pathogen Inoculation Assays: A. officinalis and A. setaceus were challenged with Phomopsis asparagi, with disease progression monitored and tissue samples collected for transcriptomic analysis [100].
• Expression Profiling: RNA-seq analysis quantified expression changes of conserved NLR genes following pathogen infection, identifying differentially expressed genes through statistical comparison of inoculated versus control plants [100].

Diagram 1: Experimental workflow for comparative NLR gene analysis, showing the integrated computational and functional approaches used to identify and characterize NLR genes across Asparagus species.

Molecular Consequences of NLR Contraction in Cultivated Asparagus

Impaired Defense Signaling in Domesticated Genotypes

Pathogen inoculation assays revealed stark phenotypic differences between asparagus species: A. officinalis exhibited clear susceptibility to Phomopsis asparagi infection, while A. setaceus remained largely asymptomatic [100]. This contrasting response correlates with differential NLR expression patterns—the majority of conserved NLR genes in cultivated asparagus showed either unchanged or downregulated expression following fungal challenge [100] [9]. This transcriptional inertia suggests a functional impairment of immune signaling mechanisms in the domesticated species, potentially resulting from artificial selection pressures that prioritized horticultural traits over defense capabilities.

The promoter regions of NLR genes in all three Asparagus species contain numerous cis-elements responsive to defense signals and phytohormones, indicating conserved regulatory potential [100]. However, the domesticated species appears to have compromised ability to activate these defense networks, pointing to disruptions in upstream signaling components or transcriptional regulators rather than promoter sequence loss per se [100].

Evolutionary Dynamics of NLR Repertoires

NLR genes in all three Asparagus species display chromosomal clustering patterns, consistent with observations in other plant species where NLRs often reside in dynamic genomic regions prone to duplication, recombination, and rearrangement [100]. This organizational feature facilitates rapid evolution of pathogen recognition specificities in wild species but may predispose these regions to contraction under domestication, particularly when pathogen pressure is reduced in agricultural environments [100].

The observed NLR contraction in cultivated asparagus follows a pattern documented in other crop species, where the genetic bottleneck of domestication often reduces diversity in disease resistance genes [100] [9]. This erosion of NLR diversity potentially narrows the genetic base for resistance breeding programs, highlighting the importance of wild germplasm conservation as a reservoir of resistance alleles [100].

Diagram 2: Logical relationships showing the cascade from domestication to increased disease susceptibility through NLR repertoire contraction and functional impairment.

Table 3: Key Research Reagents and Computational Tools for NLR Gene Analysis

Category	Specific Tool/Resource	Application in NLR Research	Key Features
Genomic Databases	PRGdb 4.0	NLR gene classification and reference data	Curated plant resistance gene database with classification tools [100]
	Pfam Database	Domain identification and verification	Comprehensive collection of protein domains and families [100]
Bioinformatics Tools	HMMER v3.1b2	Hidden Markov Model searches for NLR identification	Statistical rigor in domain detection [100] [28]
	OrthoFinder v2.2.7	Orthologous gene clustering across species	Gene length-normalized BLAST scores [100]
	MCScanX	Collinearity and whole-genome duplication analysis	Detection of syntenic blocks and evolutionary events [100] [28]
	TBtools v2.136	Integrative genomic data analysis and visualization	User-friendly interface for big biological data [100]
Expression Analysis	PlantCARE	Cis-element prediction in promoter regions	Identification of defense-related regulatory motifs [100]
	Trimmomatic v0.36	RNA-seq read quality control	Adaptor removal and quality filtering [28]
	Cufflinks v2.2.1	Transcript quantification and differential expression	FPKM normalization and statistical testing [28]
Experimental Resources	Phomopsis asparagi isolates	Pathogen challenge assays	Standardized inoculation for phenotypic assessment [100]
	Asparagus wild relatives germplasm	Comparative genomics and breeding resources	A. setaceus and A. kiusianus as resistance donors [100] [9]

The comparative genomic analysis between cultivated asparagus and its wild relatives provides compelling evidence that domestication has driven substantial contraction of the NLR gene repertoire, coupled with functional impairment of retained NLR genes. This genetic erosion likely underlies the enhanced disease susceptibility observed in commercial asparagus cultivation [100] [9].

These findings highlight the critical importance of wild germplasm as reservoirs of NLR diversity for crop improvement programs. The identified orthologous NLR pairs between wild and cultivated species represent prime candidates for functional validation and potential introduction into elite varieties through marker-assisted breeding [100]. Furthermore, the experimental frameworks and computational resources outlined in this guide provide a roadmap for similar investigations in other crop species, advancing our understanding of how domestication has reshaped plant immune systems and informing strategies to enhance disease resistance in cultivated plants through utilization of wild genetic resources.

Comparative genomics has revolutionized our understanding of how disease resistance (R) genes evolve and function across plant species. Synteny and orthology analysis provides a powerful framework for tracing the evolutionary history of conserved resistance loci by identifying genomic regions that originate from a common ancestral region. Among plant R genes, those containing a nucleotide-binding site (NBS) domain constitute one of the largest and most important families, playing critical roles in plant innate immunity against diverse pathogens [101] [102]. These NBS-encoding genes are further classified into distinct subclasses based on their N-terminal domains, primarily coiled-coil (CC-NBS-LRR or CNL) and Toll/interleukin-1 receptor (TNL) types, with TNL genes being almost nonexistent in monocot genomes [101] [102].

The conservation of R gene loci across species enables researchers to identify functionally important genetic elements through comparative approaches. Studies across grass species have revealed that R gene loci show high levels of synteny conservation, allowing researchers to trace their evolutionary trajectories [101]. Similarly, research in Sapindaceae species (Xanthoceras sorbifolium, Dinnocarpus longan, and Acer yangbiense) demonstrated that NBS-encoding genes are frequently distributed unevenly across chromosomes and often form tandem arrays, with fewer existing as singletons [48]. This structural organization has profound implications for how plants generate genetic diversity to counter rapidly evolving pathogens.

Methodological Framework for Synteny and Orthology Analysis

Computational Identification of NBS-Encoding Genes

The initial step in comparative analysis of R genes involves comprehensive identification of NBS-encoding genes across target genomes. The standard methodology employs Hidden Markov Models (HMM) based on conserved protein domains, particularly the NB-ARC domain (Pfam accession: PF00931) [48] [102]. The typical workflow begins with HMM searches against target genomes using established models, followed by confirmation of domain architecture through InterProScan analysis [102]. Sequences are then filtered to retain only those containing the essential NBS domain motifs (P-loop, Kinase-2, and GLPL), with the Kinase-2 motif particularly important for distinguishing between CNL and TNL types [102].

Figure 1: Experimental workflow for identifying and classifying NBS-encoding genes prior to synteny analysis.

Orthology Inference and Synteny Mapping

Once NBS-encoding genes are identified, orthology inference is performed using tools such as OrthoFinder with the DendroBLAST algorithm for orthogroup assignment [4]. Multiple sequence alignment is typically conducted using MUSCLE or MAFFT, followed by phylogenetic analysis to determine evolutionary relationships [102] [4]. For synteny analysis, progressive whole-genome alignment tools like Cactus enable high-confidence identification of syntenic regions across divergent species [21]. These tools facilitate the identification of collinear blocks where gene order and content are conserved between species, allowing researchers to distinguish between orthologs (genes diverging after speciation) and paralogs (genes diverging after duplication) [101] [21].

Additional analytical approaches include Ka/Ks analysis to identify selection pressures acting on R genes, where Ka/Ks > 1 indicates diversifying selection, Ka/Ks < 1 suggests purifying selection, and Ka/Ks ≈ 1 signifies neutral evolution [101]. Population genomics data can further reveal selection signatures through metrics like dN/dS ratios and population frequency distributions [21].

Comparative Evolutionary Patterns of NBS Genes Across Plant Families

Evolutionary Dynamics in Grass Species

Comprehensive analysis of 12 grass genomes has revealed distinct evolutionary patterns between different classes of NBS-encoding genes. R genes located in tandem duplication (TD) arrays evolve rapidly under diversifying selection, accumulating mutations that facilitate functional innovation to counter evolving pathogens [101]. In contrast, R singletons experience stronger purifying selection, maintaining sequence conservation and functional stability across species [101]. This evolutionary dichotomy represents complementary strategies for plant immunity: TD arrays generate diversity for recognizing novel pathogen effectors, while singletons preserve essential immune signaling components.

The distribution of NBS genes across grass species shows considerable variation linked to ploidy level and evolutionary history. Table 1 summarizes the distribution of NBS genes across representative plant species:

Table 1: Comparative Analysis of NBS-Encoding Genes Across Plant Species

Plant Species	Genome Type	Total NBS Genes	% of Total Genes	Main NBS Types	Evolutionary Pattern
Triticum aestivum [101]	Hexaploid	2,747	2.55%	CNL	Expansion
Oryza sativa [101]	Diploid	587	~1.5%	CNL	Contraction/Expansion
Setaria italica [101]	Diploid	535	~1.3%	CNL	Moderate conservation
Zea mays [101]	Tetraploid	306	0.35%	CNL	Contraction
Arabidopsis thaliana [101] [102]	Diploid	202	0.83%	TNL, CNL	Balanced
Xanthoceras sorbifolium [48]	Diploid	180	N/A	CNL, TNL	"First expansion then contraction"
Dinnocarpus longan [48]	Diploid	568	N/A	CNL, TNL	"Expansion-contraction-expansion"
Medicago truncatula [102]	Diploid	154	N/A	CNL	Species-specific expansion

Lineage-Specific Evolutionary Patterns

Different plant families exhibit distinctive evolutionary patterns of NBS genes shaped by their phylogenetic history and ecological pressures. In Sapindaceae species, researchers observed three distinct evolutionary patterns: X. sorbifolium showed "first expansion and then contraction," while A. yangbiense and D. longan exhibited "first expansion followed by contraction and further expansion" [48]. The stronger recent expansion in D. longan suggests it gained more genes to respond to various pathogens compared to A. yangbiense [48].

Similarly, studies across Brassicaceae, Fabaceae, and Rosaceae species revealed family-specific patterns. Fabaceae and Rosaceae species generally show "consistent expansion" of NBS genes, while Brassicaceae species typically display "first expansion and then contraction" patterns [48]. Even within the same family, significant variation can occur, as observed in Solanaceae, where pepper shows "contraction," tomato exhibits "first expansion and then contraction," and potato demonstrates "consistent expansion" [48].

Experimental Validation of Synteny-Based Resistance Loci Predictions

Functional Characterization Through Gene Silencing

Virus-induced gene silencing (VIGS) has emerged as a powerful technique for functionally validating NBS genes identified through synteny analysis. In a comprehensive study of cotton NBS genes, researchers identified 12,820 NBS-domain-containing genes across 34 plant species and grouped them into 603 orthogroups [4]. Expression profiling revealed that orthogroups OG2, OG6, and OG15 showed upregulated expression in various tissues under biotic and abiotic stresses in cotton accessions with differing susceptibility to cotton leaf curl disease (CLCuD) [4]. Most significantly, silencing of GaNBS (OG2) in resistant cotton demonstrated its crucial role in viral titer reduction, functionally validating its resistance activity [4].

Association Mapping and Selection Studies

Genome-wide association studies (GWAS) provide another approach for validating synteny-identified resistance loci. In Brassica napus, association mapping identified 13 significant SNP loci associated with resistance to different pathotypes of Plasmodiophora brassicae [103]. Among these, 9 SNPs mapped to the A-genome and 4 to the C-genome, with resistance genes located 0.04 to 0.74 Mb from the significant SNP markers [103]. This approach successfully linked genomic regions identified through comparative analysis with specific resistance phenotypes.

Selection mapping in maize populations improved for quantitative disease resistance to northern leaf blight (NLB) identified 25 SSR loci showing evidence of selection after multiple generations [104]. These selected loci were distributed across the genome, with particularly strong evidence on chromosome 8, where several selected loci co-localized with previously published NLB QTL and a race-specific resistance gene [104]. This demonstrates how selection mapping can complement synteny analysis for identifying functionally important resistance loci.

Research Toolkit for Synteny and Orthology Analysis

Table 2: Essential Research Reagents and Computational Tools for Synteny Analysis

Tool/Resource	Category	Primary Function	Application Example
HMMER [48] [102]	Bioinformatics Tool	Hidden Markov Model searches	Identifying NBS-encoding genes using NB-ARC domain
OrthoFinder [4]	Bioinformatics Tool	Orthogroup inference	Clustering NBS genes into orthologous groups
Cactus [21]	Comparative Genomics	Whole-genome alignment	High-confidence synteny identification across species
VISTA Browser [105]	Comparative Genomics	Genome alignment visualization	Examining pre-computed whole-genome alignments
NCBI Comparative Genome Viewer [106]	Comparative Genomics	Genome comparison	Comparing two genomes via assembly-assembly alignments
MEME Suite [102]	Bioinformatics Tool	Motif discovery	Identifying conserved protein motifs in NBS domains
TASSEL-GBS [103]	Genomics	SNP discovery and analysis	Genotyping by sequencing for association mapping
MEGA [101] [102]	Phylogenetics	Evolutionary analysis	Phylogenetic tree construction and evolutionary inference

Figure 2: Integrated workflow combining synteny analysis with functional validation approaches.

Synteny and orthology analysis has fundamentally advanced our understanding of how disease resistance genes evolve and function across plant lineages. The consistent finding that tandemly duplicated R genes evolve under diversifying selection while singleton R genes experience purifying selection reveals a sophisticated evolutionary strategy balancing innovation with conservation [101]. These insights are increasingly relevant for crop improvement programs, where understanding the evolutionary history of R genes facilitates more precise breeding strategies.

Future research directions will likely leverage pan-genome sequencing to capture the full diversity of R genes across entire genera, moving beyond single reference genomes. Additionally, the integration of machine learning approaches for predicting resistance functions from sequence data and synteny information shows promise for accelerating the identification of valuable R genes for crop breeding. As comparative genomics tools continue to advance, synteny and orthology analysis will remain fundamental for tracing the evolutionary origins of disease resistance and harnessing this knowledge for sustainable agriculture.

Conclusion

Comparative genomics of NBS domain genes has fundamentally advanced our understanding of plant immunity evolution, revealing dynamic gene family histories characterized by independent expansion and contraction events across plant lineages. The integration of robust bioinformatics methodologies with functional validation has enabled researchers to move beyond cataloging NBS gene diversity toward identifying key players in disease resistance pathways. Critical insights emerge from comparing resistant and susceptible genotypes, demonstrating how domestication and selection have sometimes compromised NLR repertoires while wild relatives preserve valuable resistance determinants. Future research directions should prioritize the development of unified annotation standards, enhanced machine learning applications for predicting resistance specificities, and the integration of pan-genomic approaches to capture the full spectrum of NBS gene diversity. These advances will accelerate the translation of genomic discoveries into durable disease resistance in crop species through marker-assisted breeding and precision genetic engineering, ultimately contributing to global food security.

References

• 1. Comparative genomics and bioinformatics approaches ... [https://pmc.ncbi.nlm.nih.gov/articles/PMC10929019/]
• 2. Comparative-genomic analysis reveals dynamic NLR gene ... [https://pmc.ncbi.nlm.nih.gov/articles/PMC10017999/]
• 3. Comparative Genomics of Non-TNL Disease Resistance Genes from Six Plant Species - PMC [https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5664099/]
• 4. Comparative analysis, diversification, and functional ... [https://www.nature.com/articles/s41598-024-62876-5]
• 5. Diversity, abundance, and domain architecture of plant ... [https://pmc.ncbi.nlm.nih.gov/articles/PMC11734081/]
• 6. TNL genes in peach: insights into the post-LRR domain - PMC [https://pmc.ncbi.nlm.nih.gov/articles/PMC4851768/]
• 7. Deciphering Plant NLR Genomic Evolution: Synteny-Informed Classification Unveils Insights into TNL Gene Loss - PubMed [https://pubmed.ncbi.nlm.nih.gov/39835721/]
• 8. The large repertoire of conifer NLR resistance genes includes drought responsive and highly diversified RNLs - PMC [https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6691002/]
• 9. Comparative analysis of the NLR gene family in ... [https://www.frontiersin.org/journals/plant-science/articles/10.3389/fpls.2025.1681919/full]
• 10. Genome-wide identification and comparative analyses of ... [https://www.sciencedirect.com/science/article/pii/S0304423825004315]
• 11. Identification and Characterization of NBS Resistance ... [https://pmc.ncbi.nlm.nih.gov/articles/PMC8585750/]
• 12. Genome-wide comparative analysis of NBS-encoding genes ... [https://bmcgenomics.biomedcentral.com/articles/10.1186/s12864-017-3682-x]
• 13. A compendium of genome-wide sequence reads from NBS ... [https://www.nature.com/articles/s41598-020-67848-z]
• 14. Genome-wide comparative analysis of NBS-encoding genes ... [https://bmcgenomics.biomedcentral.com/articles/10.1186/1471-2164-15-3]
• 15. Evidence-based gene models for structural and functional ... [https://biologydirect.biomedcentral.com/articles/10.1186/s13062-017-0191-4]
• 16. Evolutionary trajectory of pattern recognition receptors in ... [https://www.nature.com/articles/s41467-023-44408-3]
• 17. Genome‐wide identification and evolutionary analysis of ... [https://pmc.ncbi.nlm.nih.gov/articles/PMC12326289/]
• 18. The evolutionary patterns, expression profiles, and genetic ... [https://www.frontiersin.org/journals/plant-science/articles/10.3389/fpls.2023.1168124/full]
• 19. Whole genome sequencing reveals transcriptional and ... [https://www.nature.com/articles/s41598-025-18334-x]
• 20. Comparative genomics and bioinformatics approaches ... [https://www.frontiersin.org/journals/genetics/articles/10.3389/fgene.2024.1358134/full]
• 21. Rethinking de novo genes in plants: mechanisms ... [https://www.frontiersin.org/journals/plant-science/articles/10.3389/fpls.2025.1724832/full]
• 22. Comparative analysis of NBS domain sequences ... [https://www.sciencedirect.com/science/article/abs/pii/S1055790303002744]
• 23. Genome-wide identification and characterization of ALOG ... [https://www.frontiersin.org/journals/plant-science/articles/10.3389/fpls.2025.1690365/full]
• 24. A Genome-Wide Comparison of NB-LRR Type ... [https://pmc.ncbi.nlm.nih.gov/articles/PMC3887800/]
• 25. Genome-Wide Analysis of the NBS-LRR Gene Family and ... [https://www.mdpi.com/2311-7524/10/12/1293]
• 26. Comparative Genomics Reveals Gene Duplication and ... [https://www.mdpi.com/2311-7524/11/2/209]
• 27. Evolution of Gene Duplication in Plants - PMC - PubMed Central [https://pmc.ncbi.nlm.nih.gov/articles/PMC4972278/]
• 28. Genome-Wide Identification of NBS-LRR Family in Three ... [https://pmc.ncbi.nlm.nih.gov/articles/PMC12192410/]
• 29. Different Modes of Gene Duplication Show Divergent ... [https://www.frontiersin.org/journals/plant-science/articles/10.3389/fpls.2018.00161/full]
• 30. Phylogenetic and evolutionary analysis of NBS-encoding ... [https://www.sciencedirect.com/science/article/abs/pii/S1055790306005112]
• 31. Bryophytes hold a larger gene family space than vascular ... [https://www.nature.com/articles/s41588-025-02325-9]
• 32. Plant NLR diversity: the known unknowns of pan-NLRomes [https://pmc.ncbi.nlm.nih.gov/articles/PMC8226294/]
• 33. The Diversification of Plant NBS-LRR Defense Genes Directs ... [https://pmc.ncbi.nlm.nih.gov/articles/PMC5026261/]
• 34. NLRSeek: A reannotation–based pipeline for mining ... [https://www.sciencedirect.com/science/article/pii/S2662173825002085]
• 35. Macroevolution of NLR genes in family Fabaceae provides ... [https://www.sciencedirect.com/science/article/pii/S2667064X23001215]
• 36. Evolution patterns of NBS genes in the genus Dendrobium ... [https://bmcplantbiol.biomedcentral.com/articles/10.1186/s12870-022-03904-2]
• 37. Tracing the path from conservation to expansion ... [https://pmc.ncbi.nlm.nih.gov/articles/PMC11863892/]
• 38. Chapter 4 Phylogenetic Tree Visualization [https://yulab-smu.top/treedata-book/chapter4.html]
• 39. BIRCH Tutorial - Visualization of Phylogenetic Trees [https://home.cc.umanitoba.ca/~psgendb/tutorials/bioLegato/phylogeny/trees/trees.html]
• 40. ColorPhylo: A Color Code to Accurately Display Taxonomic ... [https://pmc.ncbi.nlm.nih.gov/articles/PMC3255521/]
• 41. A comparison of profile hidden Markov model procedures for ... [https://pmc.ncbi.nlm.nih.gov/articles/PMC140544/]
• 42. Drug target prediction through deep learning functional ... [https://www.nature.com/articles/s41467-024-46089-y]
• 43. Analysing protein complexes in plant science: insights and ... [https://pmc.ncbi.nlm.nih.gov/articles/PMC12098255/]
• 44. A long road ahead to reliable and complete medicinal plant ... [https://pmc.ncbi.nlm.nih.gov/articles/PMC11876585/]
• 45. InterPro: the protein sequence classification resource in 2025 [https://pmc.ncbi.nlm.nih.gov/articles/PMC11701551/]
• 46. The limitations of the current protein classification tools in ... [https://www.sciencedirect.com/science/article/pii/S0168165622000967]
• 47. PRGminer: harnessing deep learning for the prediction of ... [https://www.frontiersin.org/journals/plant-science/articles/10.3389/fpls.2025.1411525/full]
• 48. Distinct Evolutionary Patterns of NBS-Encoding Genes in ... [https://pmc.ncbi.nlm.nih.gov/articles/PMC7365912/]
• 49. Different orthology inference algorithms generate similar ... [https://pmc.ncbi.nlm.nih.gov/articles/PMC11788906/]
• 50. OrthoFinder: phylogenetic orthology inference for comparative ... [https://genomebiology.biomedcentral.com/articles/10.1186/s13059-019-1832-y]
• 51. Comparison of gene clustering criteria reveals intrinsic ... [https://genomebiology.biomedcentral.com/articles/10.1186/s13059-023-03089-3]
• 52. Organization and molecular evolution of a disease-resistance ... [https://bmcgenomics.biomedcentral.com/articles/10.1186/1471-2164-12-240]
• 53. Genome-wide identification and characterization of NBS ... [https://www.nature.com/articles/s41598-025-03507-5]
• 54. Overexpression of Lolium multiflorum LmMYB1 Enhances ... [https://pubmed.ncbi.nlm.nih.gov/37894960/]
• 55. A novel candidate cis-regulatory motif pair in the promoters ... [https://pmc.ncbi.nlm.nih.gov/articles/PMC3246208/]
• 56. High Basal Expression and Dual Stress Responsiveness of ... [https://pmc.ncbi.nlm.nih.gov/articles/PMC12473679/]
• 57. Evolutionary and Functional Insights into Rice Universal ... [https://pmc.ncbi.nlm.nih.gov/articles/PMC12561244/]
• 58. Functional analysis of plant disease resistance genes and ... [https://pubmed.ncbi.nlm.nih.gov/10458999/]
• 59. Functional analysis of plant disease resistance genes and ... [https://www.sciencedirect.com/science/article/pii/S1369526699800491]
• 60. Benchmark study for evaluating the quality of reference ... [https://pmc.ncbi.nlm.nih.gov/articles/PMC9988948/]
• 61. Report on Annotation Standards [https://www.earthbiogenome.org/report-on-annotation-standards]
• 62. Genome-Wide Comparative Analysis Reveals Similar Types ... [https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0121893]
• 63. Quality assessment of gene repertoire annotations with ... [https://www.nature.com/articles/s41587-024-02147-w]
• 64. GenomeQC: a quality assessment tool for genome assemblies ... [https://bmcgenomics.biomedcentral.com/articles/10.1186/s12864-020-6568-2]
• 65. Large-scale analysis of NBS domain-encoding resistance ... [https://pmc.ncbi.nlm.nih.gov/articles/PMC4171767/]
• 66. Genome-Wide Characterization and Expression Pattern ... [https://www.mdpi.com/1422-0067/26/18/9063]
• 67. Genome-Wide Identification and Mapping of NBS-Encoding ... [https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0034775]
• 68. Identification and Characterization of NBS Resistance ... [https://www.frontiersin.org/journals/plant-science/articles/10.3389/fpls.2021.758559/full]
• 69. Genome-wide comparative analysis of the nucleotide ... [https://www.frontiersin.org/journals/plant-science/articles/10.3389/fpls.2022.960723/full]
• 70. Functional characterization of NBS-LRR genes reveals an ... [https://bmcbiol.biomedcentral.com/articles/10.1186/s12915-024-01836-x]
• 71. Identification, characterization, and validation of NBS ... [https://pmc.ncbi.nlm.nih.gov/articles/PMC10318170/]
• 72. Recurrent Tandem Gene Duplication Gave Rise to Functionally Divergent Genes in Drosophila - PMC [https://pmc.ncbi.nlm.nih.gov/articles/PMC2878002/]
• 73. An Overview of Duplicated Gene Detection Methods: Why the Duplication Mechanism Has to Be Accounted for in Their Choice - PMC [https://pmc.ncbi.nlm.nih.gov/articles/PMC7565063/]
• 74. Tandem Duplication of Serpin Genes Yields Functional ... [https://pmc.ncbi.nlm.nih.gov/articles/PMC12646145/]
• 75. Detecting long tandem duplications in genomic sequences - PMC [https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3464658/]
• 76. Plant Comparative Genomics [https://link.springer.com/book/10.1007/978-1-0716-2429-6]
• 77. Whole-genome sequencing reveals novel tandem ... [https://www.nature.com/articles/s41467-019-09644-6]
• 78. Selective maintenance of Drosophila tandemly arranged duplicated genes during evolution - PMC [https://pmc.ncbi.nlm.nih.gov/articles/PMC2646280/]
• 79. PRGminer: harnessing deep learning for the prediction of ... [https://pmc.ncbi.nlm.nih.gov/articles/PMC12170542/]
• 80. A Genome-Wide Comparison of NB-LRR Type ... [https://www.sciencedirect.com/science/article/pii/S1016847823071157]
• 81. Analytical Validation of a Genomic Newborn Screening ... [https://pmc.ncbi.nlm.nih.gov/articles/PMC12551001/]
• 82. Genome-wide analysis of NBS-LRR genes in Rosaceae ... [https://www.frontiersin.org/journals/genetics/articles/10.3389/fgene.2022.1052191/full]
• 83. Transcriptome analysis reveals key defense genes associated ... [https://bmcgenomics.biomedcentral.com/articles/10.1186/s12864-025-12056-0]
• 84. Expression Data at MaizeGDB [https://www.maizegdb.org/expression]
• 85. Home < Expression Atlas < EMBL-EBI [https://www.ebi.ac.uk/gxa/home]
• 86. Gene Expression Omnibus - NCBI - NIH [https://www.ncbi.nlm.nih.gov/geo/]
• 87. Plant NBS-LRR proteins in pathogen sensing and host defense [https://pmc.ncbi.nlm.nih.gov/articles/PMC1973153/]
• 88. Transcriptome profiling of symptomatic vs. asymptomatic ... [https://pmc.ncbi.nlm.nih.gov/articles/PMC12220349/]
• 89. Dissecting resistance mechanisms to Agroathelia rolfsii in ... [https://link.springer.com/article/10.1007/s42161-025-01986-z]
• 90. NGS-based transcriptome analysis of starry flounder ... [https://www.sciencedirect.com/science/article/abs/pii/S0145305X25000990]
• 91. Genome-Wide Identification of NBS-LRR Family in Three ... [https://www.mdpi.com/2073-4425/16/6/680]
• 92. Virus-Induced Gene Silencing (VIGS) in Functional Genomics [https://www.mdpi.com/2311-7524/11/11/1297]
• 93. An optimized disease resistance gene cloning workflow for ... [https://pmc.ncbi.nlm.nih.gov/articles/PMC12119808/]
• 94. Truncated CMV2b N43 enhances virus-induced gene ... [https://plantmethods.biomedcentral.com/articles/10.1186/s13007-025-01446-w]
• 95. Systematic Analysis and Comparison of Nucleotide ... [https://pmc.ncbi.nlm.nih.gov/articles/PMC3735570/]
• 96. Genome-wide investigation of family‑1 UDP ... [https://www.sciencedirect.com/science/article/abs/pii/S0926669024014481]
• 97. Cotton CC-NBS-LRR Gene GbCNL130 Confers ... [https://www.frontiersin.org/journals/plant-science/articles/10.3389/fpls.2021.695691/full]
• 98. Genome‐wide association study discovered candidate genes ... [https://pmc.ncbi.nlm.nih.gov/articles/PMC5698051/]
• 99. Advances and prospects of genetic mapping of Verticillium wilt ... [https://jcottonres.biomedcentral.com/articles/10.1186/s42397-021-00109-0]
• 100. Comparative analysis of the NLR gene family in the genomes ... [https://pmc.ncbi.nlm.nih.gov/articles/PMC12507765/]
• 101. Comparative Genomics Analysis in Grass Species Reveals ... [https://www.nature.com/articles/s41598-019-47121-8]
• 102. Comparative Genomics of Non-TNL Disease Resistance ... [https://pmc.ncbi.nlm.nih.gov/articles/PMC5664099/]
• 103. Identification of resistance loci against new pathotypes ... [https://www.nature.com/articles/s41598-021-85836-9]
• 104. Selection Mapping of Loci for Quantitative Disease ... [https://pmc.ncbi.nlm.nih.gov/articles/PMC2535707/]
• 105. VISTA tools [https://genome.lbl.gov/vista/index.shtml]
• 106. Comparative Genome Viewer - Home - NCBI [https://www.ncbi.nlm.nih.gov/cgv]