Pigeonpea [Cajanus cajan (L.) Millsp.] is the sixth most important grain legume crop globally, with annual production exceeding 5 million tonnes (FAOSTAT, 2023). India accounts for approximately 70% of global production. Pigeonpea is a valuable source of dietary protein (20-22%) and essential amino acids (Saxena et al., 2010). The crop exhibits remarkable phenotypic diversity across agro-ecological zones (Upadhyaya et al., 2007). As a drought-tolerant, nitrogen-fixing legume, pigeonpea plays a crucial role in sustainable agriculture and food security (Varshney et al., 2012).
       
The pigeonpea draft genome (Varshney et al., 2012) and whole-genome resequencing of 292 accessions (Varshney et al., 2017) have provided unprecedented genomic resources. The intra-specific consensus genetic map has enhanced genomic frameworks (Bohra et al., 2012). New sources of disease resistance have been identified (Sharma et al., 2012). However, translating genomic information into practical breeding tools requires robust molecular markers.
       
Several molecular marker systems have been developed for pigeonpea, including simple sequence repeats (SSRs), single nucleotide polymorphisms (SNPs) and expressed sequence tag-derived SSRs (EST-SSRs). SSR markers in particular have been used to assess genetic diversity in the pigeonpea minicore collection (Muniswamy et al., 2019) and among cytoplasmic-genic male sterile, maintainer and inbred lines (Petchiammal et al., 2015). However, these markers often exhibit limitations such as moderate polymorphism rates, high development costs for SNP genotyping platforms, species-specificity limiting cross-species transferability and requirement for expensive detection equipment. These limitations underscore the need for alternative marker systems that combine high informativeness with cost-effectiveness and cross-species applicability.
       
Intron-spanning markers (ISM) exploit conserved exon sequences for primer design while targeting polymorphic intronic regions. The conserved primer sites ensure high amplification success, while variable intronic sequences provide polymorphism for genetic discrimination (Badoni et al., 2016). ISM-derived intron length polymorphism (ILP) markers offer advantages including gel-based detection without specialized equipment, co-dominant inheritance and potential cross-species transferability due to conserved flanking regions.
       
The ISM/ILP approach has proven effective in rice, where Badoni et al., (2016) developed 16,510 ILP markers achieving 80% amplification efficiency. Wang et al., (2005) established foundational methodology for genome-wide ILP development. These studies demonstrate the utility of intron-based markers for high-throughput genetic analysis.
       
Comparative genomics has revealed conserved intron architecture within the Papilionoideae subfamily. Pigeonpea, soybean and common bean share substantial synteny (Varshney et al., 2012; Schmutz et al., 2010, 2014). This conservation suggests potential for cross-species marker transferability, which warrants experimental verification.
       
Despite pigeonpea’s importance, systematic ISM resources have not been developed. This study presents the first comprehensive genome-wide computational identification of ISM candidate loci in pigeonpea, characterizes intron architecture and provides a foundational resource for future ILP marker development, genetic diversity analysis, QTL mapping and marker-assisted selection in breeding programs.

The present computational study was conducted at the Division of Genomic Resources, ICAR-NBPGR, New Delhi, India during the period from December 2025 to January 2026. This study is entirely computational in nature; experimental PCR validation is planned as future work.
       
Terminology: Throughout this manuscript, “ISM candidate loci” refers to computationally identified intron positions suitable for primer design. These are not experimentally validated markers but represent potential sites for ILP marker development pending PCR validation.
 
Genome data sources and processing
 
The pigeonpea reference genome annotation (GCF_000340665.2_C.cajan_V1.1) was obtained from NCBI RefSeq. This assembly comprises 590.4 Mb, of which 247.5 Mb is anchored to the 11 chromosomes (CcLG01-CcLG11) and the remainder lies on unplaced scaffolds; the genome size is estimated at 833 Mb (Varshney et al., 2012). Comparative data for soybean (Schmutz et al., 2010), chickpea (Varshney et al., 2013), common bean (Schmutz et al., 2014) and Medicago truncatula (Young et al., 2011) were compiled from published genome papers.
 
Computational pipeline
 
A Python-based pipeline was developed using Python 3.9 with the following packages: pandas 1.5.3 for data manipulation, BioPython 1.81 for sequence parsing and NumPy 1.24.0 for numerical operations. The pipeline operates in five stages (Fig 1): (1) Data acquisition; (2) GFF3 parsing and exon extraction; (3) Intron boundary calculation; (4) Filtering and deduplication; (5) Output generation. Scripts are available from the corresponding author on request.
 
 

Filtering and deduplication
 
Introns were filtered by size (50-5000 bp). The minimum threshold of 50 bp ensures sufficient sequence for reliable primer design and amplification. The upper limit of 5000 bp maintains PCR feasibility under standard conditions, as larger amplicons exhibit reduced efficiency (Wang et al., 2005; Badoni et al., 2016). Deduplication was performed by collapsing introns with identical genomic coordinates (chromosome, start, end) across transcript isoforms, ensuring each candidate locus represents a unique genomic position regardless of annotation complexity.
 
Validation candidate selection
 
Fifty ISM candidate loci were selected for future experimental validation based on: (i) optimal intron size (300-800 bp), (ii) location in stress-responsive genes and (iii) suitable flanking exon sequences. Primer design criteria include: melting temperature 55-65°C, GC content 40-60% and expected product size 400-1200 bp.

Genome-wide ISM candidate loci identification
 
Comprehensive analysis identified 139,873 unique, non-redundant ISM candidate loci from 23,948 protein-coding genes (Table 1). These represent computationally identified intron positions suitable for ILP marker development, pending experimental validation. The average introns per gene was 5.84, with mean intron length of 500 bp (median 228 bp), indicating a right-skewed distribution typical of eukaryotic genomes.


 
Comparative analysis across legume species
 
Comparative analysis revealed consistent intron architecture across legumes (Table 2). Pigeonpea exhibited 4.82 introns per  protein-coding gene, closely comparable to soybean (4.80) and chickpea (4.50). The coefficient of variation for intron length across species was 10.2% and pigeonpea statistics fall within one standard deviation of the legume family mean, supporting evolutionary conservation of gene structure. This conservation suggests potential for cross-species marker transferability, though experimental validation is required to confirm amplification success and polymorphism detection across species.


 
Practical applications
 
The identified ISM candidate loci have diverse potential applications in pigeonpea improvement programs. Upon experimental validation, these markers could facilitate: (i) genetic diversity assessment across germplasm collections, (ii) construction of high-density linkage maps, (iii) QTL mapping for agronomic traits including yield components, days to flowering and stress tolerance and (iv) marker-assisted selection for disease resistance and drought tolerance. The gel-based detection system makes ILP markers particularly suitable for resource-limited breeding programs without access to expensive genotyping platforms.
 
Study limitations
 
Several limitations should be acknowledged. First, the identified ISM candidate loci are computationally predicted and await experimental PCR validation to confirm amplification success, polymorphism rates and cross-species transferability. Second, marker identification is dependent on genome annotation quality; annotation errors or incomplete gene models may result in false positives or missed candidates. Third, comparative statistics for other legume species were derived from published literature rather than uniform re-analysis, limiting formal statistical comparisons. Fourth, intron polymorphism can only be assessed after experimental validation across diverse germplasm. Despite these limitations, the stringent filtering criteria and use of the latest RefSeq annotation ensure a high-confidence candidate set for future validation studies.

This study presents the first comprehensive genome-wide catalog of 139,873 unique ISM candidate loci from 23,948 intron-coding genes in pigeonpea. Nearly half (49.1%) fall within optimal PCR range, providing a ready shortlist for  ILP marker development. Comparative analysis revealed conserved intron architecture across legumes, suggesting potential for cross-species applications pending experimental verification. The practical significance of this resource lies in its potential applications for genetic diversity analysis, linkage mapping, QTL identification and marker-assisted breeding in pigeonpea improvement programs. Future work will focus on experimental PCR validation of prioritized candidates and polymorphism assessment across diverse pigeonpea germplasm.

The present study was supported by ICAR-NBPGR. The author acknowledges computational resources provided by ICAR-NBPGR.
 
Disclaimers
 
The views and conclusions expressed in this article are solely those of the author and do not necessarily represent the views of their affiliated institutions.
 
Data availability
 
The pigeonpea genome annotation is available from NCBI RefSeq (GCF_000340665.2).

The author declare no conflicts of interest regarding the publication of this article.