Identification of novel candidate genes for Ascochyta blight resistance in chickpea

doi:10.21203/rs.3.rs-4784305/v1

Download PDF

Article

Identification of novel candidate genes for Ascochyta blight resistance in chickpea

https://doi.org/10.21203/rs.3.rs-4784305/v1

This work is licensed under a CC BY 4.0 License

You are reading this latest preprint version

Ascochyta blight (AB), caused by the necrotrophic fungus Ascochyta rabiei, is a major threat to chickpea production worldwide. Resistance genes with broad-spectrum protection against virulent A. rabiei strains are required to secure chickpea yield in the US Northern Great Plains. Here we performed a genome-wide association (GWA) study to discover novel sources of genetic variation for AB resistance using a worldwide germplasm collection of 219 chickpea lines. AB resistance was evaluated 3, 9, 11, 13, and 14 days post-inoculation (dpi). Multiple GWA models revealed eight quantitative trait nucleotides (QTN) across timepoints mapped to chromosomes (Chr) 1, 3, 4, 6, and 7. Of these eight QTNs, only CM001767.1_28299946 on Chr 4 had previously been reported. A total of 153 candidate genes, including genes with roles in pathogen recognition and signaling, cell wall biosynthesis, oxidative burst, and regulation of DNA transcription, were observed surrounding QTN-targeted regions. QTN CM001766.1_36967269 on Chr 3 explained up to 33% of the variation in disease severity and was mapped to an exonic region of the pentatricopeptide repeat-containing protein At4g02750-like gene (LOC101506608). This QTN was validated across all models and timepoints. Further gene expression analysis on the QTNs identified in this study will provide insights into defense-related genes that can be further incorporated into new chickpea cultivars to minimize fungicide applications required for successful chickpea production.

Biological sciences/Genetics/Genetic association study/Genome wide association studies

Biological sciences/Genetics/Plant breeding

Ascochyta rabiei

genome-wide association studies

disease resistance

pulse crops

Chickpea (Cicer arietinum L.) is an annual legume or ‘pulse’ crop which seeds are an important source of carbohydrates, proteins, and dietary fibers for human nutrition. Rich in virtually all essential amino acids, chickpea seeds are an alternative protein substitute for vegetarians or those who cannot afford animal products ¹. Kabuli and desi are the two primary chickpea market classes cultivated around the world. Kabuli seeds have smooth surface, are light-colored, with shades from white to cream, typically large in size, and cultivated mostly in West Asia, North Africa, North America, and Europe ^1,2. In contrast, desi seeds have a reticulated surface, are dark-colored, with shades from light-brown to black, typically small in size, and cultivated mostly in Asia and Africa ^1–3. At present, more than 50 countries worldwide grow chickpeas every year. Global chickpea production reached 18 million tons in 2022, 2 million tons more than in 2021. India is the world’s largest chickpea producer followed by Australia, Turkey, Ethiopia, Russia, Myanmar, Pakistan, Mexico, Iran, and the United States (https://www.fao.org/faostat). In the US, ~ 80% of the national chickpea production is concentrated in Washington and the Northern Great Plains States of Montana, North Dakota, Nebraska, and South Dakota (https://quickstats.nass.usda.gov).

Ascochyta blight (AB), caused by the necrotrophic fungus Ascochyta rabiei (Pass.) Labr. (teleomorph: Didymella rabiei (Kovacheski) von Arx.), poses a threat to chickpea production worldwide ^4,5. The fungus attacks all aerial parts of chickpea plants at all growth stages ⁶. Initial symptoms of AB infections on leaves and pods are characterized by circular necrotic lesions with brown margins and a grey center, while lesions on stems are typically elongated ⁷. Lesioned stems girdle at later stages of infection and eventually break which leads to the dying off of the plant parts above the girdled portion ^3,7. Infection on pods also compromises seed set ³. Tissue collapse followed by plant death may occur following infection, especially during cool, humid weather ⁸. The pathogen survives in chickpea seeds, plant debris, and in the soil, all of which can serve as the primary sources of inoculum ⁴. AB infections usually start very localized and rapidly spread across entire fields through either wind, rain, or human activity ^4,9. Without proper control, AB outbreaks can cause complete yield loss, especially when susceptible cultivars are used ⁹.

The AB control in chickpea fields is best achieved when an integrated set of management practices is employed. Such AB management practices include the use of pathogen-free seeds and genetically resistant cultivars, seed treatment, burial of plant debris, crop rotation, and ultimately, fungicide applications as soon as AB foci are detected ⁴. However, successful AB control can be jeopardized by new virulent strains of A. rabiei that are constantly emerging. A. rabiei can reproduce sexually, allowing new combinations of virulence genes to appear and therefore new pathotypes that are either resistant to existing fungicide chemistries or that can overcome host defense mechanisms conferred by the current set of resistance genes in use ³. Numerous A. rabiei pathotypes and physiological races have already been described in major chickpea production areas ^10,11. Historically, genetic resistance to AB has been associated with kabuli chickpea types ². Current resistant chickpea cultivars, however, only provide moderate resistance to the known A. rabiei pathotypes, which has not been sufficient to prevent economic loss due to AB outbreaks in the Northern Great Plains ^5,12. In addition, available fungicides registered to control AB have been losing efficacy over the years ⁵. Therefore, a continuous search for new resistance genes is encouraged to secure chickpea production.

This study utilized genome-wide association (GWA) mapping to identify new source of genomic variation underlying AB resistance using a global collection of chickpea lines. AB resistance was evaluated at multiple time points across this study, and multiple GWA models identified significant associations across timepoints. Candidate genes and their possible role in chickpea resistance mechanisms against A. rabiei are elucidated.

2.1. Genome Coverage

A total of 2,425 high-quality Single Nucleotide Polymorphism (SNP) markers were used to perform GWA mapping. SNP markers were randomly distributed across the chickpea genome and their location in the CDC Frontier reference genome ASM33114v1 is depicted in Fig. 1a. SNPs are located at the beginning, middle, and far end of all eight chickpea chromosomes. Marker density per chromosome (Chr) ranged from 103 observed for Chr 8 to 758 observed for Chr 4. Average distance between markers varied from 0.06 megabase pair (Mbp) on Chr 4 to 0.27 Mbp on Chr 5. SNPs were as close as 1 bp from one another and as far as 7,449,018 bp on Chr 2 (Fig. 1a).

Additional insight into genome coverage and mapping resolution was gained by studying linkage disequilibrium (LD) decay rates within chromosomes (Supplementary Fig. 1a-h). LD decay rate varied depending on the chromosome. The fastest rate was observed in Chr 2 (r² = 0.2; 0.18 Mbp) while the slowest rate was observed in Chr 7 (r² = 0.2; 2.40 Mbp). Genome-wide LD decay at r² = 0.2 was 0.57 Mbp indicating that there should be one SNP for every 0.57 Mbp or less (Fig. 1b). Genome-wide LD decay was higher than the average marker distance, suggesting satisfactory mapping resolution. With a genome size of 530.8 Mbp, 930 evenly distributed SNPs (530.8/0.57) are required for good genome coverage.

2.2. Population Structure

Evidence of population structure within the chickpea genotype panel was suggested by both principal component (PC) and hierarchical clustering analyses. SNP-based PC biplots indicated the presence of three to four major clusters (Fig. 2a) although cluster number could be higher since PC 1, 2, and 3 explained only 33.79% of the total genotypic variation. The k-means statistic indicated five as the optimal number of clusters (BIC = 1097.788). Clusters 1, 2, and 3 accommodated 47, 65, and 86 of the 219 chickpea lines used in GWA mapping while clusters 4 and 5 accommodated only 11 and 10 lines, respectively (Fig. 2b). Clustering techniques also provided evidence for genotypic diversity within the cultivated chickpea pool. Cultivated lines were scattered across positive and negative quadrants of PC1, PC2, and PC3 (Fig. 2a) and clustered in all five groups (Fig. 2b). The two wild accessions used in this study appeared in the negative quadrants of PC1 and PC2, and clustered together in group 1, suggesting that they are closely related. Chickpea lines came from all over the world (Fig. 2c). Thirty-eight percent of the lines originated from India, the world’s largest chickpea producer. Countries such as Iran and Pakistan were also well represented, accounting for 22% and 7% of the lines, respectively. Indian lines spread across positive and negative quadrants of PC1 and PC2 and clustered in groups 1, 2, 3 suggesting genetic diversity even amongst lines from the same country of origin. Likewise, Iranian lines clustered in all five groups.

2.3. Disease Severity

Circular necrotic lesions on leaves and stems are the typical symptoms of A. rabiei infection in chickpeas (Fig. 3). Disease progress was monitored by recording AB severity scores 3, 9, 11, 13, and 14 days post-inoculation (dpi). Visual scores in the x-axis range from 1 (no symptoms) to 9 (dead plant) (Fig. 4a). At the early time point of 3 dpi, the fungus was still in its biotrophic phase and none of the plants showed AB symptoms (Fig. 4a). With the switch from the biotrophic to the necrotrophic stage around 9 dpi, early-stage symptoms manifested on leaves and stems and increased in severity over time as shown in Fig. 4a. Disease ratings at 9, 11, 13, and 14 dpi ranged from 1 to 4, 1 to 4, 1 to 6, and 1 to 8, respectively (Fig. 4a).

Wald’s test for fixed effects revealed significant genotype effect for disease score among the 220 USDA lines and the four checks on days 11, 13, and 14 dpi (p < 2.2e^− 16). Genotype effect was less prominent at 9 dpi with a p-value of 0.06. As expected, kabuli genotypes had lower severity scores than desi genotypes. Overall, heritability estimates of disease severity scores were relatively low and ranged from 0.14 at 11 dpi to 0.27 at 13 dpi. Figure 4b displays BLUP-based genotype raking for AB resistance. Average BLUP severity scores of bottom-five genotypes (most susceptible) were twice that of top-five resistant genotypes. The resistant checks ND Crown and CDC Frontier had their resistance confirmed as they occupied the 8th and 18th places in the ranking, respectively. The susceptible check Sierra ranked amongst susceptible lines whereas the susceptible check Sawyer was intermediately ranked (Fig. 4b).

2.4. Genome-Wide Association Mapping

To discover novel genetic variation underlying AB resistance in this chickpea panel, we performed GWA mapping using a single-locus model approach (MLM), and three multi-locus model approaches (MLMM, FarmCPU, and BLINK) available in the GAPIT package ¹³. GWA mapping was also performed across all time points to elucidate candidate genes involved in both early and late stages of infection. Table 1 contains significant SNPs by model across all time points. Manhattan as well as quantile-quantile (Q-Q) plots for SNP-trait associations over time detected by different models are presented in Fig. 5 as well as Supplementary Figs. 2, 3, and 4. Eight QTN were identified as being potentially involved in AB regulation: 2 each on chromosomes 1, 3, and 4 plus one each on chromosomes 6 and 7. The multi-locus models MLMM, FarmCPU, and BLINK identified QTNs for AB severity even at early stages of disease infection such as 9 and 11 dpi. At 9 dpi, BLINK identified SNP CM001766.1_36967269 located at 36967269 bp on Chr 3 (-log₁₀ p-value = 7.10) and SNP CM001769.1_34406910 located at 34406910 bp on Chr 6 (-log₁₀ p-value = 6.61) explaining 21.5% and 35.5% of the total phenotypic variation, respectively (Table 1). SNP CM001766.1_36967269 on Chr 3 was validated by FarmCPU (-log₁₀ p-value = 3.76) while SNP CM001769.1_34406910 on Chr 6 was validated by MLMM (-log₁₀ p-value = 3.81). As for 11 dpi, FarmCPU and BLINK models showed identical results with SNPs mapped to Chr 1 (CM001764.1_10036603; -log₁₀ p-value = 3.66) and Chr 3 (CM001766.1_36967269; -log₁₀ p-value = 4.09). In contrast to multi-locus models, MLM only identified significant variants at 13 dpi. The largest number of SNPs were mapped at day 13 by Farm-CPU on chromosomes 1 (CM001764.1_5724993), 3 (CM001766.1_36967269), 4 (CM001767.1_20036993), and 7 (CM001770.1_14470637). The SNP CM001766.1_36967269 on Chr 3 was validated by all four models with transformed p-values located above the Bonferroni threshold (Table 1). Such QTN could potentially explain up to 33% of the total variation in AB severity scores (Table 1). BLINK results for day 13 suggest that two QTNs alone, located on Chr 3 and 4, explained 68.2% of the total phenotypic variation. QTN on Chr 3 was evidenced by all three multi-locus models on day 14 (Table 1). Additionally, QTN on Chr 3 was identified by both FarmCPU and BLINK across all timepoints.

Table 1

Genome-wide association mapping results across models and time points.
Timepoint	Chr	SNP	SNP Position (bp)	Model	-log₁₀(p-value)	PVE (%)
9 dpi	3	CM001766.1_36967269	36967269	FarmCPU	3.76	NA
	3	CM001766.1_36967269	36967269	BLINK	7.10	21.5
	6	CM001769.1_34406910	34406910	MLMM	3.81	NA
	6	CM001769.1_34406910	34406910	BLINK	6.61	35.5
11 dpi	1	CM001764.1_10036603	10036603	FarmCPU	3.66	NA
	1	CM001764.1_10036603	10036603	BLINK	3.66	NA
	3	CM001766.1_36967269	36967269	FarmCPU	4.09	NA
	3	CM001766.1_36967269	36967269	BLINK	4.09	NA
13 dpi	1	CM001764.1_5724993	5724993	FarmCPU	4.10	NA
	3	CM001766.1_36263134	36263134	MLM	3.99	NA
		CM001766.1_36967269	36967269	MLM	5.07	31.9
				MLMM	5.51	1.8
				FarmCPU	13.35	1.8
				BLINK	10.06	33.0
	4	CM001767.1_20036993	20036993	FarmCPU	3.68	NA
		CM001767.1_28299946	28299946	MLMM	3.75	NA
		CM001767.1_28299946	28299946	BLINK	5.40	35.2
	7	CM001770.1_14470637	14470637	FarmCPU	4.62	NA
14 dpi	3	CM001766.1_36967269	36967269	MLMM	3.66	NA
				FarmCPU	3.73	NA
				BLINK	3.73	NA

Chr = Chromosome; SNP = Single nucleotide polymorphism; PVE = Percentage of phenotypic variance explained; NA = Not applicable (PVE was only retrieved for those SNPs who exceed the Bonferroni threshold).

2.5. Candidate Genes

A candidate gene search revealed a series of genes surrounding the eight QTNs identified in the GWAS analysis. A total of 153 genes located within a 200-kb window (100 kb upstream and downstream) of significant SNPs are reported in Supplementary Table 1. Six of the eight significant SNPs are inside genes (Supplementary Table 1). Of these eight SNPs, four are inside or flanking uncharacterized genes. The significant SNPs CM001764.1_10036603, and CM001764.1_5724993 on Chr 1, CM001766.1_36967269 on Chr 3, CM001767.1_28299946 on Chr 4, CM001769.1_34406910 on Chr 6, and CM001770.1_14470637 on Chr 7 belong to the genes MMS19 nucleotide excision repair protein homolog (LOC101495813), chitin elicitor receptor kinase 1-like (LOC101510407), pentatricopeptide repeat-containing protein At4g02750-like (LOC101506608), probable inorganic phosphate transporter 1–3 (LOC101497071), uncharacterized LOC113787172 (LOC113787172), and uncharacterized LOC101512131 (LOC101512131), respectively (Supplementary Table 1). The SNPs CM001766.1_36263134 on Chr 3 and CM001767.1_20036993 on Chr 4 are outside coding regions. BTB/POZ domain-containing protein At1g03010-like (LOC101502431) and uncharacterized LOC105851914 (LOC105851914) at 142 bp and 2989 bp away from these two SNPs are the closest genes (Supplementary Table 1).

Gene families with putative role in plant defense mechanisms against fungal infections including pathogen recognition and signaling (chitin elicitor receptor kinase 1-like, probable leucine-rich repeat receptor-like protein kinase At1g35710, LysM domain receptor-like kinase 3) ^14–16, hormone-mediated signal transduction (transcription factor MYBS3, transcription factor MYBS3-like, ethylene-responsive transcription factor 1B-like), cytoplasm oxidative burst (monothiol glutaredoxin-S11, monothiol glutaredoxin-S6, monothiol glutaredoxin-S6-like, glutaredoxin-C11) ¹⁴, and cell wall biosynthesis (long-chain acyl-CoA synthetase 1 isoform X, cytochrome P450 84A1-like)¹⁷ are among the candidate genes reported within the search window (Supplementary Table 1).

The SNP CM001766.1_36967269 on Chr 3 validated across all models and timepoints (Table 1) fell into an exonic region of the pentatricopeptide repeat-containing protein At4g02750-like gene (LOC101506608). Two other genes, the long chain acyl-CoA synthetase 1 isoform X1 gene (LOC101505979) involved with cell wall biosynthesis, and the WRKY transcription factor 23 gene (LOC101495104) involved with regulation of DNA-templated transcription, were also found near this SNP (Fig. 6a). Among the twenty-six candidate genes reported within the 200-kb search window, LOC101506608, LOC101505979, and LOC101495104 are most likely responsible for regulation of resistance against A. rabiei in QTN CM001766.1_36967269. The average disease score of GG genotypes for this SNP was 24 to 72% higher than that of CC genotypes across trials and time points (Fig. 6b,c).

Ascochyta blight is an intricate disease that affects stems, leaves, seed set, and ultimately yield of chickpeas. Usually, growers adopt an integrated approach to manage AB in chickpea fields involving the use of pathogen-free seeds, genetically resistant cultivars, and fungicide applications ⁴. However, effective control of AB in the face of new virulent strains that can resist existing fungicides and overcome host resistance genes can be challenging. Prioritizing resistant cultivars over chemical control is a cost-effective and environmentally sustainable alternative ¹⁸. Current resistance genes, however, do not provide enough protection to prevent economic losses, and chemical applications are still required ^5,12. Identifying new, powerful resistance genes to combat AB ensures yield while avoiding environmental contamination of chickpea fields. This study mapped new genomic regions associated with AB resistance among chickpea lines adapted to cultivation in major chickpea-producing areas. Candidate genes with putative roles in plant defense responses against biotic and abiotic stresses were also reported. Additional transcriptome and proteome studies are needed to further pinpoint resistance genes that could be later incorporated into elite breeding lines.

3.1. Genetic variation among chickpea lines and breeding for Ascochyta blight resistance

Disease histograms for disease-specific symptom development at 3, 9, 11, 13, and 14 dpi demonstrate the progression of symptoms over time. At 14 dpi, disease severity scores were normally distributed, a pattern commonly observed in disease screening trials ^9,19. A normal distribution pattern means that most of the lines are moderately resistant/susceptible to A. rabiei while just a few of them are either highly resistant or highly susceptible. Significant genotype effect identified in the analysis of variance and heritability estimates suggests that part of the variation in disease ratings is attributed to genetic causes. Additionally, PC analysis and hierarchical clustering based on SNP data further demonstrate that these chickpea lines differ from one another at the DNA level. Genotypic variation for the trait of interest is required for genetic improvement. Positioning of checks in the disease ranking agreed with what was expected according to their known resistance level except for Sawyer. Seven lines had lower disease severity scores than the best-ranked resistant check (ND Crown) suggesting that we can develop cultivars with even better resistance than what is currently available in the market.

3.2. Genome-Wide Association Models and Quantitative Trait Nucleotide Discovery

The GWA models used to map QTN for AB resistance likely differed in terms of QTN detection due to how each of them controls for false positive associations. MLM usually controls false positives well by fitting Q and K ²⁰. However, MLM assumes that only a single locus is behind trait expression which may be incorrect in the case of AB resistance in chickpea as numerous small-effect loci have already been reported ^9,19,21–23. MLMM assumes the involvement of multiple loci, and in addition to Q and K, it also adds associated markers known as Pseudo QTNs to improve power ²⁴. Confounding effects between K and marker matrices in both MLM and MLMM models, however, often result in loss of statistical power which is the reason why FarmCPU and BLINK models tend to map more QTNs ²⁵. The models we used are ordered as follows in terms of QTN detection power: BLINK > FarmCPU > MLMM > MLM ^24–26. As expected, we identified more QTNs across timepoints with the multi-locus models FarmCPU and BLINK than with the multi-locus model MLMM and the single-locus model MLM. Observed p-values in MLM and MLMM Q-Q plots tended to be slightly lower than the expectation compared to FarmCPU and BLINK (Fig. 5 and Supplementary Figs. 2–4) which can be an indication of overfitting and consequently loss of true associations in these models. Nevertheless, using MLM and MLMM is encouraged to validate strong signals found in more sophisticated models such as FarmCPU and BLINK. In total, eight QTNs for AB resistance were found considering all models and timepoints. The significant SNP CM001767.1_28299946 on Chr 4 matched the QTL mapped by Carmona et al. ²⁷ and Singh et al. ²⁸. The remaining QTNs have not as yet been reported to the best of our knowledge.

4.3. Candidate genes and their role in host defense mechanisms

During infection, pathogens secrete a plethora of bioactive molecules called effectors and enzymes to degrade cell wall components into the host. These effectors often have specific targets in the host plant that they manipulate to aid infection. As a necrotrophic pathogen, A. rabiei relies on killing of host cells for nutrient supply. Necrotrophic fungi secrete effectors to hijack innate plant defense responses and thereby evoking cell death which will result in the release of nutrients that are taken up by the fungus ^29–35. To combat A. rabiei, chickpea genotypes make use of their sophisticated immune system consisting of physical barriers, pathogen recognition receptors, oxidative burst and signal transduction cascades to switch on genes encoding pathogenesis-related proteins ^14,36,37. Our candidate gene search revealed genes potentially involved in all these plant defense mechanisms (Supplementary Table 1).

A gene involved with cutin/wax biosynthesis (long chain acyl-CoA synthetase 1 isoform X1 - LOC101505979) was found 9,263 bp away from the SNP CM001766.1_36967269 on Chr 3. A lignin-biosynthesis gene (cytochrome P450 84A1-like - LOC101512776) was found 55,180 bp away from the SNP CM001770.1_14470637 on Chr 7. The cell wall is the first line of defense against pathogens as it acts as a physical barrier preventing pathogen entry. Under the initial stages of A. rabiei infection, resistant chickpea genotypes up-regulate genes involved with lignin, cutin, and wax biosynthesis which results in cell wall thickening and reinforcement on pathogen penetration sites ³⁷.

The two QTNs on Chr 1 appear to be involved with pathogen recognition. The SNP CM001764.1_10036603 is 11,252 bp away from a LysM domain receptor-like kinase 3 (LOC101496137) and the SNP CM001764.1_5724993 is within a chitin elicitor receptor kinase 1-like (LOC101510407). Protein receptor-like kinases embedded in plant cell membranes recognize microbial or modified-self ligand invasion patterns and send signals to activate downstream defense responses ³⁶. LOC101510407 plays a role in chickpea resistance against fungal infections. Yadav et al. (2023) found a gene encoding a chitin elicitor receptor kinase 1-like to be upregulated in the xylem of a resistant chickpea line while downregulated in a susceptible line under initial stages of Fusarium oxysporum infection.

After successful pathogen recognition, the chickpea immune system induces oxidative stress to kill or damage pathogen hyphae ³⁶. We believe the QTN CM001766.1_36263134 mapped on Chr 3 may be involved with AB resistance as the genes monothiol glutaredoxin-S11 (LOC101500507), glutaredoxin-C11 (LOC101500829), monothiol glutaredoxin-S6 (LOC101501469), and monothiol glutaredoxin-S6-like (LOC101501787) were found within its 200-kb search window, all of which are involved in inducing oxidative burst. Overexpression of LOC101500507, LOC101500829, and LOC101501469 was observed in a chickpea-resistant line as a defense response to fungal attack ¹⁴.

Candidate genes encoding transcription factors (TF) or other proteins involved in DNA-templated transcription were found on chromosomes 1, 3, 4, and 7. TFs are proteins that, in response to biotic and abiotic stresses, regulate gene expression by binding to promoters of target genes ³⁸. TFs are downstream activated by jasmonic acid, abscisic acid, and ethylene signaling pathways^37,39. Ethylene-induced TFs transcription factor MYBS3-like (LOC101496138), and transcription factor MYBS3 (LOC101494932) were located at a close distance from SNP CM001764.1_10036603 on Chr 1, and an ethylene-responsive transcription factor 1B-like (LOC101496212) was located at a close distance from SNP CM001767.1_28299946 on Chr 4. The candidate gene WRKY transcription factor 23 (LOC101495104) located 35,271 bp away from CM001766.1_36967269 on Chr 3 was up-regulated in a drought-tolerant chickpea genotype under drought stress ⁴⁰.

Pentatricopeptide repeat-containing (PPR) proteins have previously been reported as have been associated with resistance to fungal diseases in plants ^9,37,41–43. Sahin et al.⁹ identified a PPR protein involved with AB resistance on the chickpea chromosome 1. Gene expression profiling during A. rabiei infection showed up-regulation of a PPR protein exclusively in resistant chickpea genotypes ³⁷. Since the SNP at 36,967,269 bp on Chr 3 is located within the PPR protein At4g02750-like gene, we suspect that the base substitution (C > G) inactivated the gene which resulted in increased disease severity levels in GG chickpea lines. Transcriptome and proteome analyses could be used further to decipher the involvement of PPR protein At4g02750-like as well as other candidate genes in chickpea defense mechanisms against A. rabiei.

4.1. Greenhouse Trials

To identify the underlying genomic regions involved in chickpea resistance to AB, two disease trials were performed at the Dalrymple Research Greenhouse complex in Fargo, ND, United States. Two hundred and twenty lines from the USDA chickpea core collection plus the AB resistant checks CDC Frontier and ND Crown, and susceptible checks Sawyer and Sierra were screened for AB severity after A. rabiei inoculation. Chickpea lines consisted of cultivars (15), landraces (4), cultivated (194), breeding (3), and wild (2) materials. Desi chickpeas accounted for 70% of the lines, whereas kabuli chickpeas accounted for 23%. Two and fourteen (~ 7%) lines had their improvement status and seed type uncertain, respectively. While the chickpea lines hailed from across the globe, the majority of them came from India, Iran, and Pakistan, countries that figure among the world’s top-ten chickpea producers.

Trials were laid out in a 20x12 un-replicated diagonal arrangement format. Single USDA line plots consisting of four plants side by side were randomly assigned to the 20 rows and 12 columns. Check plots were replicated five times each and assigned to rows and columns in a diagonal fashion ⁴⁴. Chickpea seeds were sown in plastic cones (1.5” d x 8.25” h) (Stuewe & Songs, Inc., OR, USA) (1 plant per cone) filled with potting mix Pro-mix PGX (Premier Tech Horticulture, Quakertown). Water was applied to plants as needed. Fertilization was performed twice and consisted of applying a 20:20:20 NPK soluble fertilizer solution to plants (2 tsp per gallon) two weeks after sowing and one day before inoculation. Room conditions for chickpea growth were set to 65 to 70 °F and 16h-light 8h-dark cycles.

4.2. Disease Screening

Chickpea varieties were inoculated with the A. rabiei APNS4 strain. For spore production, A. rabiei APNS4 cultures were maintained on chickpea agar medium under 24-h light at 22 ± 2°C. Chickpea agar medium was prepared by adding 100 g of chickpea seeds of cultivar Sawyer into 1,000 ml of boiling distilled water. After 30 minutes, the chickpea seeds were strained out and 15 g of Agar was added to the medium. The chickpea agar medium was autoclaved and poured into petri plates. A. rabiei APNS4 conidia were harvested by adding 0.01% Tween 20 (Acros Organics, NJ, USA) to the 2-week-old A. rabiei APNS4 plates and gentle scraping of the fungal culture using a sterile L-shaped glass spreader to release the spores. The spore suspension was filtered through a double layer of Miracloth (EMD Millipore, Burlington, MA, USA) and adjusted the concentration to 3 × 10⁵ conidia/ml ⁴⁵ Four 18-day-old chickpea plants per variety were uniformly inoculated using a paint-spray gun (Anest Iwata-Medea Inc, Portland, OR). Additionally, four plants of each check (Sierra, Sawyer, CDC Frontier, and ND Crown) were inoculated with MQ water to serve as water control to compare symptom development with that of inoculated plant checks. The inoculated plants were incubated in humidity chambers (Phytotronic Inc., Earth City, MO) at > 95 %RH and 22 ± 2°C for 48 h and then transferred back to the greenhouse room where they were kept on the benches. Water control checks were kept away from the pathogen-inoculated plants. The disease severity was checked for each plant individually at 3, 9, 11, 13, and 14 dpi using a modified scale of 1–9 (Table 2) as described by Kaur et al. (2013).

Table 2

Modified disease scale used to quantify Ascochyta blight severity in the chickpea plants.
Rating	Symptoms	Resistant Class
1	No Symptoms	Asymptomatic
2	Minute lesions prominent on Apical Stem	Resistant
3	Lesions visible on < 10% of all leaves and slight stem girdling	Resistant
4	Lesions visible on 10–25% of all leaves and stem girdling on < 10% of plant.	Moderately Resistant
5	Lesions on 25% of all leaves, and stem girdling on 10% of plant	Moderately Resistant
6	Lesions on 50% of all leaves, defoliation – broken, breaking and drying of branches due to stem girdling – slight to moderate	Moderately Susceptible
7	Lesions present on most of the plants, stem girdling on 50% of the plant	Moderately Susceptible
8	Lesions and stem girdling as in 7, but 50% of the plant killed	Susceptible
9	Lesions profuse on all parts of the plant, stem girdling on more than 50% of the plant and death of most plants	Susceptible

4.3. DNA Extraction and Genotyping

Genotypic data was obtained from leaf DNA samples of young chickpea seedlings. The DNeasy Plant Mini Kit (Qiagen) was used to extract DNA from fresh and/or lyophilized leaf tissue following the manufacturer’s protocol. Total DNA was eluted in 100 µL of Buffer AE and quantified using the Qubit dsDNA BR Assay kit and Qubit 4.0 fluorometer (Life Technologies Corporation). DNA solution was concentrated to 25 ng/µL of DNA and samples were shipped to HudsonAlpha Institute for Biotechnology for genotyping-by-sequencing (GBS).

Restriction enzyme ApeKI ⁴⁷ was used to prepare dual-indexed GBS libraries which were later combined into a single pool and sequenced across 1.5 lanes of a NovaSeq S1 x 100-pb run. About ≈ 1000 M pass filter reads were generated in the sequencing step. Reads with quality scores ≥Q30 were aligned to the chickpea reference genome ASM33114v1.

4.4. Marker Filtering and Imputation

Single polymorphism nucleotide (SNP) markers produced by GBS were aligned to the eight chromosomes of the chickpea reference genome. The initial marker matrix contained the 220 USDA chickpea lines and 5,363 SNP markers. Marker filtering steps adopted were MAF ≥ 0.05, heterozygosity ≤ 0.10, and individual and marker call rate ≥ 0.80. The final cleaned marker matrix contained 219 individuals and 2,425 high-quality SNPs that met the filtering criteria. Missing values (NAs), which corresponded to only 2.52% of the total SNPs, were set to the mean value for each particular SNP and rounded to the nearest integer. All marker filtering and imputation steps were performed using the ASRgenomics R package ⁴⁸.

4.5. Linkage Disequilibrium

Squared correlation coefficients (r²) for all pair-wise marker combinations were estimated as a measure of linkage disequilibrium. Non-linear models were fitted for r² values from pair-wise SNP combinations and their respective physical distances using Hill and Weir’s recombination formula⁴⁹ implemented in the nls() function ⁵⁰. These r² values were plotted against physical distances for visualization (Fig. 1b, Supplementary Fig. 1a-h).

4.6. Principal Component Analysis and Hierarchical Clustering

Principal component (PC) and hierarchical clustering analyses were performed using data from the 2,425 high-quality SNPs to investigate the genetic relationship among the 219 chickpea lines that formed the mapping population. PCs are often included as cofactors in GWA models to control for population structure and avoid false marker-trait associations ⁵¹. prcomp() and ggplot() functions of base R and ggplot2^50,52 were used to estimate and plot genotype coordinates for PCs 1, 2, and 3, respectively. To compute the distance matrix required for hierarchical clustering we used the dist() function with the argument method = “euclidian” ⁵⁰. The function find.clusters() implemented in the adegenet package⁵³ was used to find the optimal number of clusters. This function runs successive k-means with an increasing number of clusters (k) and reports Bayesian Information Criterion (BIC) values for each k. We considered the optimal number of clusters that k after which drops in BIC values were minimal. The hclust() function⁵⁰ was then used to cluster genotypes based on Ward’s method described in ⁵⁴.

4.7. Phenotypic Analyses and Best Linear Unbiased Prediction of AB Severity Scores

All statistical analyses were conducted in the R statistical software Version 4.3.2 ⁵⁰. Wald’s test for fixed effects was used to investigate whether genotype effects were significant ⁵⁵. The asreml() function implemented in the ASReml package⁵⁵ was used to fit a linear mixed model and extract single Best Linear Unbiased Prediction (BLUP) values of mean AB severity scores for genotypes at each timepoint separately. Trial was considered as fixed, while genotype and genotype-by-trial interaction were fitted as random effects in the model. We also included the first-order autoregressive process [AR (1)] into the error structure to accommodate spatial trends across the columns and rows of each trial. Overall heritability of disease severity scores was estimated by the Cullis method ⁵⁶.

4.8. Genome-Wide Association Mapping

GWA mapping was conducted across all timepoints using the single-locus mixed linear model (MLM) ⁵⁷, and the multi-locus models multiple loci mixed model (MLMM) ²⁴, fixed and random model circulating probability unification (FarmCPU) ²⁵, and Bayesian-information and linkage-disequilibrium iteratively nested keyway (BLINK)²⁶ implemented in the R package GAPIT ¹³. These GWA models differ in terms of computational efficiency, statistical power, and control of false positives ⁵⁸. The first three principal components were considered in the models to account for population structure in our dataset. The single-locus model MLM assumes that only a single-locus is responsible for the phenotypic variation observed while MLMM, FarmCPU, and BLINK assume that more loci can be involved ⁵⁹. MLM introduces population structure (Q) and Kinship (K) among individuals as covariates in the model to reduce false marker-trait associations ⁵⁷. MLMM goes a step further and adds associated markers named Pseudo QTNs as covariates to the model while keeping Q and K matrices unchanged ^25,26. FarmCPU eliminates confounding effects between kinship and tested markers by fitting a fixed effect model with Pseudo QTNs iteratively with a random effect model responsible for estimating the Pseudo QTNs ²⁵. The Pseudo-QTNs in FarmCPU are derived through restricted maximum likelihood (REML) and then used to derive the kinship matrix ²⁵. BLINK replaces the REML approach in FarmCPU with the Bayesian Information Criterion and uses linkage disequilibrium information to break the biased assumption that causal genes are equally distributed across the genome ²⁶. For more details about MLM, MLMM, FarmCPU, and BLINK reference Zhu et al. (2008), Segura et al. (2012), Liu et al. (2016), and Huang et al. (2019). Those SNPs whose transformed p-values were greater than the comparison-wise error rate threshold proposed by Li and Ji⁶⁰ and described in Bari et al.⁶¹ were considered to be significant and were visualized in Manhattan plots (Fig. 5; Supplementary Fig. 2–4). To calculate the Li and Ji threshold ⁶⁰, we first estimated the effective number of independent tests (Meff) from the correlation matrix and eigenvalue decomposition of 2,425 SNPs. We then used the equation \(\:{\alpha\:}_{P}=1{-(1-{\alpha\:}_{e})}^{1/Meff}\), where \(\:{\alpha\:}_{P}\) is the computed comparison-wise error rate, and \(\:{\alpha\:}_{e}\) is the trial-wise error rate here defined as 0.05 to calculate the threshold. Comparison-wise error rate in this study was \(\:-{\text{log}}_{10}\left({\alpha\:}_{p}\right)=\:3.65.\) The Bonferroni threshold \(\:{-log}_{10}\:(0.05/\text{2,425})=4.68\) was also presented in the Manhattan plots for comparison. In GAPIT, the percentage of variance explained (PVE) by QTNs is only retrieved for those SNPs who exceed the Bonferroni threshold ¹³. Available PVEs are shown in Table 2.

4.9. Candidate Gene Search

A search window of 200-kb, 100 kb upstream and 100 kb downstream of each QTN, was defined for candidate gene search ⁹. Information about gene annotation was retrieved from the National Center for Biotechnology Information (NCBI) database (https://www.ncbi.nlm.nih.gov/datasets/taxonomy/3827/).

We identified a total of eight candidate QTNs regulating AB resistance in our chickpea population on chromosomes 1, 3, 4, 6, and 7. The multi-locus models BLINK and FarmCPU detected more QTNs than the multi-locus model MLMM and the single-locus model MLM. The QTN CM001766.1_36967269 on Chr 3 was mapped across all timepoints (9, 11, 13, and 14 dpi) by BLINK and FarmCPU, and across all models (MLM, MLMM, FarmCPU, and BLINK) at 13 dpi. This QTN alone explained up to 33% of the variation in disease severity scores. A total of 153 candidate genes were mapped within a 200-kb window from QTNs. Genes implicated in plant defense mechanisms against biotic and abiotic stresses including cell wall thickening, pathogen perception, oxidative burst, signal transduction, and regulation of DNA transcription are among the candidate genes identified in this study. The QTN on Chr 3 falls into an exonic region of the pentatricopeptide repeat-containing protein At4g02750-like gene (LOC101506608). Two other genes, the long chain acyl-CoA synthetase 1 isoform X1 gene (LOC101505979) involved with cell wall biosynthesis, and the WRKY transcription factor 23 gene (LOC101495104) involved with regulation of DNA-templated transcription, are robust candidates to be associated with AB regulation in this genomic region. Further functional studies around these eight QTN-targeted regions will help us to pinpoint resistance genes for later use in breeding.

Funding

Funding for this work was made possible by North Dakota Department of Agriculture through the Specialty Crop Block Grant Program (22–237). We acknowledge the support from USDA-NIFA Hatch Project ND01513 for Nonoy Bandillo and ND02249 for Malaika Ebert. This work was also supported by the USDA National Institute of Food and Agriculture, Crop Protection and Pest Management Program through the North Central IPM Center (2022-70006-38001). This work used resources of the Center for Computationally Assisted Science and Technology (CCAST) at North Dakota State University, Fargo, ND, USA which were made possible in part by NSF MRI Award No. 2019077.

Competing interests

Authors declare that they have no conflict of interest.

Author Contribution

N.B., M.K.E., P.F., C.C., and K.M., provided the resources and conceived the manuscript. F.D.D., A.A., M.M., H.N., R.S., L.P., H.W., and G.R. performed the greenhouse trials and collected the data. F.D.D. and S.A.A. analyzed the data. F.D.D. and A.A. wrote the initial draft which was later edited by all authors. All authors read and agreed to the published version of the manuscript.

Acknowledgement

The authors thank Julie Hochhalter and her team for their support during greenhouse trials.

Data Availability

The datasets generated and analysed during the current study are available in the CyVerse repository, https://data.cyverse.org/dav-anon/iplant/home/francoisedariva/gwas_ascochytablight_paper.

Jukanti, A. K., Gaur, P. M., Gowda, C. L. L. & Chibbar, R. N. Nutritional quality and health benefits of chickpea (Cicer arietinum L.): a review. British Journal of Nutrition 108, 11–26 (2012).
Purushothaman, R., Upadhyaya, H. D., Gaur, P. M., Gowda, C. L. L. & Krishnamurthy, L. Kabuli and desi chickpeas differ in their requirement for reproductive duration. Field Crops Res 163, 24–31 (2014).
Pande, S. et al. Ascochyta blight of chickpea (Cicer arietinum L.): a review of biology, pathogenicity, and disease management. Aust J Agric Res 56, 317 (2005).
Gan, Y. T., Siddique, K. H. M., MacLeod, W. J. & Jayakumar, P. Management options for minimizing the damage by Ascochyta blight (Ascochyta rabiei) in chickpea (Cicer arietinum L.). Field Crops Res 97, 121–134 (2006).
Wise, K. A., Bradley, C. A., Pasche, J. S. & Gudmestad, N. C. Resistance to QoI fungicides in Ascochyta rabiei from chickpea in the Northern Great Plains. Plant Dis 93, 528–536 (2009).
Manjunatha, L. et al. Present status on variability and management of Ascochyta rabiei infecting chickpea. Indian Phytopathol 71, 9–24 (2018).
Sharma, M. & Ghosh, R. An update on genetic resistance of chickpea to Ascochyta blight. Agronomy 6, 18 (2016).
Khaliq, I., Moore, K. & Sparks, A. H. The relationship between natural rain intensity and Ascochyta blight in chickpea development. Eur J Plant Pathol 164, 313–323 (2022).
Şahin, E. S., Talapov, T., Ateş, D., Can, C. & Tanyolaç, M. B. Genome wide association study of genes controlling resistance to Didymella rabiei Pathotype IV through genotyping by sequencing in chickpeas (Cicer arietinum). Genomics 115, 110699 (2023).
Chongo, G., Gossen, B. D., Buchwaldt, L., Adhikari, T. & Rimmer, S. R. Genetic diversity of Ascochyta rabiei in Canada. Plant Dis 88, 4–10 (2004).
Manjunatha, L. et al. Diversity Profiling and the distribution of pathotypes and mating types in Indian isolates of Ascochyta rabiei causing Ascochyta blight of chickpea. PhytoFrontiers 3, 569–581 (2023).
Wise, K. A. et al. Sensitivity of Ascochyta rabiei populations to prothioconazole and thiabendazole. Crop Protection 30, 1000–1005 (2011).
Wang, J. & Zhang, Z. GAPIT Version 3: Boosting power and accuracy for genomic association and prediction. Genomics Proteomics Bioinformatics 19, 629–640 (2021).
Yadav, P. et al. Comprehensive transcriptome analyses of Fusarium-infected root xylem tissues to decipher genes involved in chickpea wilt resistance. 3 Biotech 13, 390 (2023).
Soltabayeva, A. et al. Receptor-like kinases (LRR-RLKs) in response of plants to biotic and abiotic stresses. Plants 11, 2660 (2022).
Buendia, L., Girardin, A., Wang, T., Cottret, L. & Lefebvre, B. LysM receptor-like kinase and LysM receptor-like protein families: an update on phylogeny and functional characterization. Front Plant Sci 9, (2018).
Wang, X., Kong, L., Zhi, P. & Chang, C. Update on cuticular wax biosynthesis and its roles in plant disease resistance. Int J Mol Sci 21, 5514 (2020).
Copati, M. G. F., Pessoa, H. P., Dariva, F. D., Castro Filho, M. N. de & Nick, C. Tomato families possessing resistance to late blight also display high-quality fruit. Acta Sci Agron 46, e66790 (2024).
Raman, R. et al. Genome-wide association analyses track genomic regions for resistance to Ascochyta rabiei in Australian chickpea breeding germplasm. Front Plant Sci 13, 877266 (2022).
Price, A. L. et al. Principal components analysis corrects for stratification in genome-wide association studies. Nat Genet 38, 904–909 (2006).
Ilyas, A., Mirza, S. A., Hussain, K., Ilyas, A. & Shahzad, K. A comprehensive review on genetic resistance of chickpea to Ascochyta blight. Journal of Plant Pathology 104, 1337–1354 (2022).
Farahani, S. et al. Genome-wide association mapping for isolate-specific resistance to Ascochyta rabiei in chickpea (Cicer arietinum L.). Physiol Mol Plant Pathol 121, 101883 (2022).
Alo, F. et al. Novel genomic regions linked to Ascochyta blight resistance in two differentially resistant cultivars of chickpea. Front Plant Sci 13, 762002 (2022).
Segura, V. et al. An efficient multi-locus mixed-model approach for genome-wide association studies in structured populations. Nat Genet 44, 825–830 (2012).
Liu, X., Huang, M., Fan, B., Buckler, E. S. & Zhang, Z. Iterative usage of fixed and random effect models for powerful and efficient genome-wide association studies. PLoS Genet 12, e1005767 (2016).
Huang, M., Liu, X., Zhou, Y., Summers, R. M. & Zhang, Z. BLINK: a package for the next level of genome-wide association studies with both individuals and markers in the millions. Gigascience 8, giy154 (2019).
Carmona, A., Castro, P., Perez-Rial, A. & Die, J. V. Genomic data of two chickpea populations sharing a potential Ascochyta blight resistance region. Data Brief 50, 109624 (2023).
Singh, R., Kumar, K., Purayannur, S. & Verma, P. K. Genomics-assisted genetics of complex regions from chickpea chromosome 4 reveals two candidate genes for Ascochyta blight resistance. Plant Science 334, 111781 (2023).
Friesen, T. l., Chu, C., Xu, S. S. & Faris, J. D. SnTox5–Snn5: a novel Stagonospora nodorum effector–wheat gene interaction and its relationship with the SnToxA –Tsn1 and SnTox3–Snn3 –B1 interactions. Mol Plant Pathol 13, 1101–1109 (2012).
Friesen, T. L. et al. Host-selective toxins produced by Stagonospora nodorum confer disease susceptibility in adult wheat plants under field conditions. Theoretical and Applied Genetics 118, 1489–1497 (2009).
Friesen, T. L., Zhang, Z., Solomon, P. S., Oliver, R. P. & Faris, J. D. Characterization of the interaction of a novel Stagonospora nodorum host-selective toxin with a wheat susceptibility gene. Plant Physiol 146, 323–324 (2008).
Friesen, T. L., Meinhardt, S. W. & Faris, J. D. The Stagonospora nodorum -wheat pathosystem involves multiple proteinaceous host‐selective toxins and corresponding host sensitivity genes that interact in an inverse gene‐for‐gene manner. The Plant Journal 51, 681–692 (2007).
Friesen, T. L. et al. Emergence of a new disease as a result of interspecific virulence gene transfer. Nat Genet 38, 953–956 (2006).
Liu, Z. et al. The Tsn1 –ToxA interaction in the wheat–Stagonospora nodorum pathosystem parallels that of the wheat–tan spot system. Genome 49, 1265–1273 (2006).
Liu, Z. H. et al. Genetic and physical mapping of a gene conditioning sensitivity in wheat to a partially purified host-selective toxin produced by Stagonospora nodorum. Phytopathology 94, 1056–1060 (2004).
Jha, U. C., Sharma, K. D., Nayyar, H., Parida, S. K. & Siddique, K. H. M. Breeding and genomics interventions for developing Ascochyta blight resistant grain legumes. Int J Mol Sci 23, 2217 (2022).
Singh, R. et al. A global transcriptome and co-expression analysis reveals robust host defense pathway reprogramming and identifies key regulators of early phases of Cicer-Ascochyta interactions. Molecular Plant-Microbe Interactions 35, 1034–1047 (2022).
Alves, M. et al. Transcription factor functional protein-protein interactions in plant defense responses. Proteomes 2, 85–106 (2014).
Berrocal-Lobo, M. & Molina, A. Ethylene response factor 1 mediates Arabidopsis resistance to the soilborne fungus Fusarium oxysporum. Molecular Plant-Microbe Interactions 17, 763–770 (2004).
Badhan, S., Kole, P., Ball, A. & Mantri, N. RNA sequencing of leaf tissues from two contrasting chickpea genotypes reveals mechanisms for drought tolerance. Plant Physiology and Biochemistry 129, 295–304 (2018).
Favre, F. et al. First Vanilla planifolia High-density genetic linkage map provides quantitative trait loci for resistance to Fusarium oxysporum. Plant Dis 107, 2997–3006 (2023).
Bardak, A., Çelik, S., Erdoğan, O., Ekinci, R. & Dumlupinar, Z. Association mapping of Verticillium wilt disease in a worldwide collection of cotton (Gossypium hirsutum L.). Plants 10, 306 (2021).
Park, S. et al. Proteome analysis of disease resistance against Ralstonia solanacearum in potato cultivar CT206-10. Plant Pathol J 32, 25–32 (2016).
Murillo, D. & Gezan, S. FielDHub: A Shiny App for Design of Experiments in Life Sciences. <https://CRAN.R-project.org/package=FielDHub> (2024).
Harveson, R. M. et al. Ascochyta blight of chickpeas. Plant Health Prog 12, 30 (2011).
Kaur, L. et al. Combining Ascochyta blight and Botrytis grey mould resistance in chickpea through interspecific hybridization. Phytopathol Mediterr 52, 157–165 (2013).
Elshire, R. J. et al. A robust, simple genotyping-by-sequencing (GBS) approach for high diversity species. PLoS One 6, e19379 (2011).
Gezan, S., de O. A. A., G. G., and M. D. ASRgenomics: An R package with complementary genomic functions. (2022).
Hill, W. G. & Weir, B. S. Variances and covariances of squared linkage disequilibria in finite populations. Theor Popul Biol 33, 54–78 (1988).
R Core Team. R: A language and environment for statistical computing. (2023).
Wen, J. et al. Association mapping of major economic traits and exploration of elite alleles in Prunus sibirica. Euphytica 219, 39 (2023).
Hadley Wickham. ggplot2: Elegant graphics for data analysis. (Springer-Verlag, New York, 2016).
Jombart, T. adegenet: a R package for the multivariate analysis of genetic markers. Bioinformatics 24, 1403–1405 (2008).
Ferreira, L. & Hitchcock, D. B. A comparison of hierarchical methods for clustering functional data. Commun Stat Simul Comput 38, 1925–1949 (2009).
The VSNi Team. asreml: fits linear mixed models using REML. (2023).
Cullis, B. R., Smith, A. B. & Coombes, N. E. On the design of early generation variety trials with correlated data. J Agric Biol Environ Stat 11, 381–393 (2006).
Zhu, X., Li, S., Cooper, R. S. & Elston, R. C. A unified association analysis approach for family and unrelated samples correcting for stratification. The American Journal of Human Genetics 82, 352–365 (2008).
Xiao, Y., Liu, H., Wu, L., Warburton, M. & Yan, J. Genome-wide association studies in maize: praise and stargaze. Mol Plant 10, 359–374 (2017).
Soto-Cerda, B. J., Aravena, G. & Cloutier, S. Genetic dissection of flowering time in flax (Linum usitatissimum L.) through single- and multi-locus genome-wide association studies. Molecular Genetics and Genomics 296, 877–891 (2021).
Li, J. & Ji, L. Adjusting multiple testing in multilocus analyses using the eigenvalues of a correlation matrix. Heredity (Edinb) 95, 221–227 (2005).
Bari, Md. A. Al et al. A greenhouse-based high‐throughput phenotyping platform for identification and genetic dissection of resistance to Aphanomyces root rot in field pea. The Plant Phenome Journal 6, e20063 (2023).

No competing interests reported.

Download PDF

Editorial decision: Revision requested
04 Oct, 2024
Reviews received at journal
18 Sep, 2024
Reviews received at journal
14 Sep, 2024
Reviewers agreed at journal
29 Aug, 2024
Reviewers agreed at journal
29 Aug, 2024
Reviewers invited by journal
27 Aug, 2024
Editor assigned by journal
27 Aug, 2024
Editor invited by journal
21 Aug, 2024
Submission checks completed at journal
13 Aug, 2024
First submitted to journal
22 Jul, 2024

You are reading this latest preprint version

Identification of novel candidate genes for Ascochyta blight resistance in chickpea

Status:

Version 1

Abstract

Figures

1. Introduction

2. Results

2.1. Genome Coverage

2.2. Population Structure

2.3. Disease Severity

2.4. Genome-Wide Association Mapping

2.5. Candidate Genes

3. Discussion

3.1. Genetic variation among chickpea lines and breeding for Ascochyta blight resistance

3.2. Genome-Wide Association Models and Quantitative Trait Nucleotide Discovery

4.3. Candidate genes and their role in host defense mechanisms

4. Material and methods

4.1. Greenhouse Trials

4.2. Disease Screening

4.3. DNA Extraction and Genotyping

4.4. Marker Filtering and Imputation

4.5. Linkage Disequilibrium

4.6. Principal Component Analysis and Hierarchical Clustering

4.7. Phenotypic Analyses and Best Linear Unbiased Prediction of AB Severity Scores

4.8. Genome-Wide Association Mapping

4.9. Candidate Gene Search

5. Conclusions

Declarations

Funding

Competing interests

Author Contribution

Acknowledgement

Data Availability

References

Additional Declarations

Supplementary Files

Status:

Version 1