Genome‐wide association study of variation in cooking time among common bean (Phaseolus vulgaris L.) accessions using Diversity Arrays Technology markers

Stored grains of common bean (Phaseolus vulgaris L.) develop the hard‐to‐cook trait (HTC), which is manifested in a prolonged cooking time, thereby imposing time and energy constraints. The objective of this study was to determine variation in cooking time among common bean genotypes and to identify single nucleotide polymorphism (SNP) markers associated with cooking time.

The common bean plays a critical role in the nutrition security of a large segment of the world population, especially in third world countries.Its dry grains are used as a major source of dietary protein.
Cooking is a fundamental part of bean preparation, and it inactivates anti-nutritive factors, increases digestibility, and improves the sensorial quality of beans (Costa et al., 2006).Some samples of dry beans have been found to require a long cooking time, which is time-and energy-consuming (Bressani & Chon, 1996).The cooking time of beans has been reported to be influenced by a diversity of factors, including genetic differences, growth environment, postharvest handling, storage conditions such as temperature and humidity, storage time, and treatments before cooking (Arruda et al., 2012).
Farmers and traders commonly store the grains for long periods before availing them to end users.When storage occurs in adverse conditions of high temperature and high relative humidity, the grains develop the hard-to-cook (HTC) phenomenon, which increases the cooking time of beans.In addition, the improperly stored grains become discolored and with decreased water absorption capacity (Ousman et al., 2013).Further, Vindiola et al. (1986) reported that although all bean samples were affected by the HTC phenomenon, the rate of hardening differed among varieties.
The formation of insoluble pectates at the cell wall and the middle lamella that renders the tissue more resistant to cell separation during cooking is believed to be the cause of HTC (Hentges et al., 1990;Shomer et al., 1990).Pectin is made up of complex acid polysaccharides with a backbone of galacturonic acid residue with an alpha-1,-4-glycosidic linkage (Atkinson et al., 2002).The increase of phenolic compounds that cause lignification of cells has also been proposed to cause HTC (Garcia et al., 1998).A dual-enzyme mechanism has been proposed to explain the development of HTC conditions in storage at elevated temperatures and relative humidity (Jones & Boulter, 1983).This theory suggests that at high temperatures and relative humidity, the pectin methylesterase (PME) enzyme hydrolyzes pectin molecules in the middle lamella, forming pectic acid and methanol.Concurrently, the phytase enzyme hydrolyzes phytic acid in cotyledon cells to release inorganic phosphate and magnesium ions.The magnesium and calcium ions released in the cells migrate to the middle lamella and produce an insoluble magnesium pectinate and calcium pectinate that cement cells together hardening the cell wall (Jones & Boulter, 1983).This hypothesis was supported by the presence of more water-soluble pectin (8.44 mg/g) in varieties with shorter cooking time than slow cooking beans (5.51 mg/g) (Njoroge et al., 2014).
The use of deoxyribonucleic acid (DNA) analysis techniques in common bean breeding programs has improved our understanding of genetic factors controlling various traits.The genome-wide association study (GWAS) is a popular method used to identify quantitative trait loci (QTL) associated with bean traits.GWASs are mainly concerned with determining alleles associated with various single nucleotide polymorphisms (SNPs) and making statistical comparisons to identify SNPs associated with a particular trait (Resende et al., 2018).
Previous studies have identified QTLs that control cooking time on Chromosomes 2, 3, and 6 using GWAS (Cichy et al., 2015) and on Chromosomes 1, 2, 3, 5, 6, 10, and 11 using QTL mapping (Berry et al., 2020).The identified regions require further exploration to determine their robustness and stability across different genetic backgrounds, growth environments, and storage conditions.This study aimed to assess common bean accessions for variation in cooking time and to identify SNP markers associated with cooking time.GLPx92, and GLP1192a).The accessions belonged to different markets/seed classes, including small whites, blacks, yellows, cariocas, brown and tan, medium whites, purples, large reds, small reds, calima, sugars, and pintos (Table 1).Seeds from a single plant randomly selected from each accession were multiplied during the rainy season of the year 2019 (September 2019-January 2020) at the Jomo Kenyatta University early June (Long rains) and again between October and November (short rains) with a monthly mean of 142 and 116 mm, respectively.The mean annual maximum and minimum temperatures were 26 C and 15 C, respectively.Data recorded during seed multiplication in the field included days to flowering, days to maturity, number of pods per plant, number of seeds per pod, seed weight, and plot seed yield as described by Van Schoonhoven and Pastor-Corrales (1987).Harvesting was done manually by picking pods of each accession; the harvested pods were then sundried in a screenhouse for 2 weeks to dry.Threshing of dry pods and sorting of the seeds were also done manually.

| Incubation of seed and determination of cooking time
The cooking time was determined on each accession using freshly harvested seeds and aged seeds.The aging process involved storing the seeds in a thermostatically controlled incubator (HP300 G, Wincom, Hunan Sheng, China) at a temperature of 35 C and 50% relative humidity for a period of 4 months.Freshly harvested seeds removed from the incubator after aging treatment were stored at À20 C to prevent further aging during the cooking experiment, which took a period of 2 months.A sample of 100 whole grain seeds of each accession was rinsed and soaked in distilled water for 16 h.Hard-shelled seeds that failed to absorb water due to impermeable seed coats were sorted out and excluded from the cooking experiment.Soaked seeds were then subjected to standard cooking using distilled water at 96 C in a thermostatically controlled water bath (WBU 45;Memmert,Schwabach,Germany).A total of 10 seeds were sampled from the cooking water bath after 20 min and at 5 min intervals thereafter without interrupting the boiling.The 10 cooked bean seeds were cooled in cold water for a minute, and their softness/hardness was determined using a subjective finger-pressing method (Kinyanjui et al., 2015;Vindiola et al., 1986).
The bean seeds were considered cooked when the cotyledon disintegrated on pressing and felt soft (lack of graininess).Cooking of the beans was done in duplicate by two people, and the percentage of cooked beans in a batch was expressed as a function of time.Briefly, the procedure involved germinating the seeds and growing the young seedlings in vermiculite for a period of 10 days.The Nucleomag ® plant DNA extraction kit (Macherey-Nagel AG, Switzerland) was used for DNA isolation from young leaves' tissue of each common bean accession.The quality and quantity of DNA were visualized using gel electrophoresis on a 0.8% agarose.Genotyping by sequencing (GBS) using Diversity Arrays Technology Sequencing (DArTseq™) technology was used to identify variability in SNP markers.DNA libraries were constructed according to DArTseq complexity reduction method through the digestion of genomic DNA using a combination of two restriction enzymes (PstI and MseI) and ligation of barcoded adapters followed by polymerase chain reaction (PCR) amplification of adapter-ligated fragments.Libraries were sequenced using single-read sequencing runs for 77 cycles (Kilian et al., 2012).Illumina Hiseq2500 platform was used for highthroughput sequencing, and scoring of markers was achieved using DArTsoft14 Software Version 1.5.2.beta (Diversity Arrays Technology 2017) as SilicoDArT markers and biallelic SNP markers.Markers were scored as binary for presence/absence (1 and 0, respectively) of  the restriction fragment with the marker sequence in the genomic representation of the sample.Both SilicoDArT markers and SNP markers were aligned to the reference genomes of the common bean (Phaseolus vulgaris 442 version 2.1) with a blast alignment of minimum 90% sequence identity and expected value of 5 Â 10 À10 to identify chromosome positions (Goodstein et al., 2012).

| Data analysis
Logistic regression modeling was used to describe the relations between cooking time and the percentage of cooked bean seeds as described by Wafula et al. (2020).Cooking time was defined as the time corresponding to the probability that 95% of the bean seeds would be cooked.The intercept and regression coefficients were used to generate graphs of cooking time for different common bean accessions.Analysis of variance was performed on the obtained cooking time using R software (version 4.0.2), and the mean values of different accessions were compared using the least square difference (LSD) at 0.05% significance level.Complete clustering analysis was conducted using NbClust package of the R software considering Euclidean distances to determine the genetic relationship among genotypes.

| Linkage disequilibrium (LD)
A total of 19,188 SNP markers with minor allele frequency (MAF) > 0.05 and integrity >0.8 were selected and used for subsequent analysis.Principal component (PC) (Q matrix) and the relative kinship (K matrix) analyses were conducted using VanRaden method within R-based GAPIT package version 0.3.4 in order to account for the population structure.The first three PCs were used to construct the PC matrix.LD was estimated between SNPs on each chromosome by calculating the square value of correlation (r 2 ) between the pair of markers using KDCompute software 0.6.1.

| Marker-trait association
Based on the 19,188 SNP markers and cooking time of 194 fresh and 222 aged common bean accessions, association analysis was performed using genome-wide association mapping to identify SNPs associated with cooking time.Marker-trait association analysis was conducted using the generalized linear model (GLM) and compressed mixed linear models (CMLM) of the default settings of Genomic Association and Prediction Integrated Tool (GAPIT) software version 0.3.4.
via the KDCompute interface (https://kdcompute.seqart.net/kdcompute/login).The GWAS threshold for the significant markertrait association was the F-test for testing the null hypothesis that there is no association between the SNP and the trait.

| Identification of potential candidate gene
Potential candidate genes that were flagged by the SNPs significantly associated with cooking time were identified using the JBrowse search tool against the reference genome of the common bean (P.vulgaris 442 version 2.1), which is available on the Phytozome database (www.phytozome.net)(Goodstein et al., 2012).The maximum threshold for identification of the potential candidate genes was set at 100 kb around the position of the trait-associated SNPs.
Additionally, the significant SNP sequences were utilized in a basic local alignment search tool for nucleotides (BLASTn) within the National Center for Biotechnology Information (NCBI) database to find potential matches (www.ncbi.nlm.nih.gov).

| RESULTS
The results show that cooking time for fresh and aged seeds varied among common bean accessions.The cooking time of fresh seeds ranged from 28.1 to 72.2 min with a mean of 40.8 min, whereas that of aged seeds ranged from 32.1 to 96.3 min with a mean of 54.9 min (Table 1 and Supporting Information Appendix Table 1).Storage of the seeds increased cooking time of the accessions, but the increase varied among common bean accessions.Accessions NUA (Ciankui) had the least increase in cooking time due to storage of 0.3% in comparison to commercial variety GLP2 (rosecoco) with the least increase in cooking time of 0.8% among the commercial varieties (Table 2).The cooking time profile of fresh and aged seeds was sigmoid; the mean difference between the cooking time of fresh and aged seeds was due to the prolonged lag phase followed by the slightly less steep exponential phase observed in the cooking time of aged seeds (Supporting Information Appendix Figure 1).
Commercial varieties evaluated in this study were categorized into two groups based on the least significant differences in cooking time; GLP2 and GLP24 had a shorter cooking time, whereas GLPx92 and GLP1127a had a longer cooking duration.GLP2 had the shortest cooking duration of 47.2 min, whereas GLP1127a had a longer cooking time of 61.1 min for aged seeds.Among the bean accessions evaluated, GBK034996 had the shortest cooking time (32.1 min), whereas GBK035370 had the highest cooking time of 96.3 min for aged seeds (Table 1).A total of 133 and 65 bean accessions had a shorter cooking time for fresh and aged seeds, respectively, than commercial variety GLP24 (Supporting Information Appendix Table 1).
There was a significant moderate positive correlation (0.59) between the cooking time of fresh and aged seeds.Cooking time for both fresh and aged seeds had a significant extremely weak negative association with days to maturity (À0.2), whereas cooking time of aged seeds had a significant but weak negative correlation with the number of pods per plant (À0.18), pod length (À0.18), and seed weight (À0.25) (Table 3).
Cluster analysis based on selected agronomic traits and average cooking time grouped the 194 genotypes into two major groups.The largest group constituted 77.3% of the genotypes, which had the highest seed weight and duration to maturity with a mean of 45.5 g and 82.1 days, respectively, and had the shortest cooking time of 44.2 min.On the other hand, the second group constituted 22.6% of the genotypes with higher cooking time (59.2 min), lower seed weight (26.5 g), and shorter duration to maturity (81.3 days) (Figure 1).
The common bean accessions used in this study belonged to 12 seed classes commonly found in the eastern Africa region.The results show significant differences in cooking time for fresh and aged seeds among seed classes.The The carioca seed class had the shortest cooking time, whereas medium whites had the longest cooking time for both fresh and aged seeds (Supporting Information Appendix -Figure 2).The seed classes were categorized into seven groups based on the least significant difference in cooking duration (Supporting Information Appendix Table 2).The profile of cooking time for fresh and aged seeds for small and medium whites was more distinct from the rest due to their prolonged cooking time mainly at the lag phase of the curve (Supporting Information Appendix Figure 3 and 4).markers on Chromosome 11.In the PC analysis, the first PCs accounted for 53.8% of the variation, whereas the second and third PCs accounted for only 6.9% of the variation (Figure 2).LD analysis conducted using 1988 loci pairs showed pairwise 917,276 loci with an average r 2 of 0.43 and ranged from 0.36 on Chromosome 11 to 0.53 on Chromosome 1 and extended to an average distance of 4406.2 kb.About 2.9% of SNPs were in complete LD (r 2 = 1).The average D 0 was 0.88 ranging from 0.84 on Chromosome 11 to 0.91 on Chromosome 1 (Figure 3).Genome-wide analysis revealed two significant SNP markers associated with the cooking time of aged seeds on Chromosome 10 (Figure 4).However, no significant SNP markers were identified associated with the cooking time of fresh seeds.The most significant SNP marker was 100,096,770jFj0-21:G > A-21:G > A on Chromosome 10 at location 5,600,323 with a p value of 6.9 Â 10-7, which explained 36% of the phenotypic variation.The second SNP marker that was significantly associated with the cooking time of aged seeds was 3,377,419jFj0-24:A > T-24:A > T in Chromosome 10 at location 4,468,450 with a p value of 9.8 Â 10-6, which explained 34% of the phenotypic variation.The two significant SNPs 100,096,770jFj0-21:G > A-21:G > A and 3,377,419jFj0-24: A > T-24:A > T had an allelic effect of 13.21% and 10.8%, respectively, and were in strong LD with r 2 of 0.79, D prime of 0.95 (Supporting Information Appendix Table 3).The region around these SNP markers also had several other SNP markers near the significance threshold with a p value ranging from 3.02 Â 10-5 to 6.52 Â 10-5 due to LD (Figure 2 and Supporting Information Appendix Table 3). in the NCBI database revealed five genes with no defined function (Table 4).

| DISCUSSION
The variation in cooking time was higher for aged seeds when compared to fresh seeds.However, the cooking time for fresh seeds among bean accessions was significantly different.This indicates that progress can be made when selecting common bean accessions with a shorter cooking time using both fresh and aged seeds because high narrow sense heritability of 0.76 and 0.74 for cooking time has been reported in previous studies by Elia (2003) and Jacinto-Hernandez et al. (2003), respectively.The significant difference observed between the cooking time of aged and fresh seeds indicates that the storage conditions significantly affected this trait.It is also evident that the extent of increase in cooking time due to storage varied among common bean accessions.This suggests that common bean susceptibility to hardening during storage varies among accessions.
Accessions NUA (Ciankui) and commercial variety GLP2 (rosecoco) are less susceptible to aging due to storage in adverse conditions.Previous studies have reported an increase in the cooking time of common beans stored in higher temperatures and relative humidity as demonstrated in this study (Nyakuni et al., 2008;Ousman et al., 2013).Nyakuni et al. (2008) reported an increase in cooking time of four varieties stored in ambient temperatures and relative humidity of 63%-74%, and the percentage increase in the cooking time varied among varieties.Several other studies have reported that some varieties have a shorter cooking time, whereas others have a prolonged cooking time.(Bressani & Chon, 1996;Kinyanjui et al., 2015;Nyakuni et al., 2008).
The cooking time for commercial varieties observed in this study follows the same trend as reported in previous studies, where rosecoco (GLP2) and red haricot (GLP24) have a relatively shorter cooking time in comparison to pinto (GLPx92) and Canadian Wonder (GLP1127a) Kinyanjui et al., 2015.The genetic diversity in cooking time among the common bean accessions evaluated provides variability that can be utilized in breeding.The accessions with shorter cooking time can serve as genetic resources for selection or hybridization schemes to generate new common bean varieties that are easy to cook.Accessions with shorter cooking time can be improved through direct selection based on other desirable traits.
The initial lag phase of the cooking time curve creates much of the cooking time difference between fresh and aged seeds.The lag phase may be the stage at which the pectin is solubilized within the middle lamella to allow water to imbibe into the cells of the cotyledon.Njoroge et al. (2014) reported that varieties that have a shorter cooking time had a higher hot water-soluble pectin (8.44 mg/g) than slow cooking beans like pinto (5.51 mg/g).If this is the case, the modification of the composition of the middle lamella may make beans easy to cook as proposed by Broughton et al. (2003).reported that cooking time is sigmoid with the lag and exponential phase being influenced by variety and storage.
The moderate (0.59) correlation in cooking time between fresh and aged seeds indicates that accessions with a higher cooking time of fresh seeds are more susceptible to hardening during storage in adverse conditions.This suggests that freshly harvested seeds can be used to an extent for indirect selection for easy-to-cook accessions.
The significant extremely weak negative association of cooking time with duration to maturity, pod length, and seed weight indicates that common bean accessions that have a longer duration to mature, have longer pods, and have larger seeds tend to have a slightly shorter cooking time.Similarly, common bean accession with a higher number of pods per plant will tend to have a slightly longer cooking time for aged seeds.This could be attributed to LD where some genes tend to be inherited together or the presence of a third variable such as the prevailing temperature and humidity during growth.Similar negative correlation results between cooking time, duration to maturity (À0.44), and seed weight (À0.21) were reported in a previous study (Cichy et al., 2019).A study conducted by Berry et al. (2020) also revealed a negative correlation between cooking time and seed weight, which ranged from À0.3 to À0.8, depending on the environment the common bean was grown in.
Classification of common bean market classes is based on seed size and color.Seed color is determined by the presence and concentrations of flavanol glycosides, anthocyanins, and condensed tannins in the seed coat (Reynoso et al., 2006).Lei et al. (2020) classified those that weigh less than 25 g per 100 seeds as small-seeded, whereas those that range from 25 to 40 g as medium-sized, and those that weigh more than 40 g are classified as large-seeded.The cluster analysis results in this study classified the accession into two groups mostly based on seed size.The results also indicate that large-seeded accessions had a relatively shorter cooking time, unlike the mediumand small-seeded genotypes.Seed size depends on the genetic difference among varieties and can be traced back to the origin of the common bean, namely, Mesoamerican and Andean gene pools (Angioi et al., 2010).The Andean gene pool is generally large-seeded, whereas the Mesoamerican gene pool is small-seeded and adapted to lower altitudes and higher temperatures (Beebe et al., 2011).
The significant results for the cooking time differences among the common bean seed classes used in this study confirm that the trait varies between and within seed classes as reported by Cichy et al. (2015).However, the findings in this study contradict those of Cichy et al. (2015), which found the white seed class to have the shortest cooking time in relation to other bean classes.This could be due to differences in the genetic background of the white bean accessions used in these two studies, the environment they were grown in, and the interaction between the genotypes and the environment.In this study, the common bean accessions were sourced from the National Gene Bank of Kenya, which had been collected mainly from Kenya, whereas Cichy et al. (2015) evaluated common bean accessions of Andean origin collected from Africa and North America.Therefore, this study supports the theory of the formation of insoluble pectin as the cause of the hard-to-cook trait.The difference in cooking time of stored bean accession is due to the activity of galacturan 1,4-alpha-galacturonidase and polygalacturonase enzymes that hydrolyze pectin.The hydrolyses of pectin could probably have happened during storage as proposed by Jones and Boulter (1983) for pectin to translocate to the middle lamella and combine with magnesium and calcium ions to produce an insoluble magnesium pectinate and calcium pectinate that cement cells together, hardening the cell wall.Alternatively, the hydrolyses of pectin could have occurred during the presoaking treatment of aged seeds before cooking.
Pectin is made up of complex acid polysaccharides with a backbone of galacturonic acid residue with an alpha-1,4-glycosidic linkage.
Homogalacturonan-rich pectin is commonly found in the middle lamellar region of plant cell walls where two cells border (Atkinson et al., 2002).Galacturan 1,4-alpha-galacturonidase enzyme is known to hydrolyze the first group of glycosidic bonds from the nonreducing end of the substrate, whereas polygalacturonase enzymes break down the pectin components found in the middle lamella of plant cells after PME makes the polymeric backbone accessible.The combined effect of both PME and pectinase enzymes has been reported to give softer fruits and vegetables at the end of maturation (Phutela et al., 2005).
The second most significant SNP associated with the cooking time marker may be due to LD (r 2 = 0.74, D 0 = 0.95) because of its proximity to the gene that controls HTC.The genes identified around this SNP allele variant marker may not be involved in the control of cooking time but based on LD can be used to determine the haplotype containing genes of interest.The identification of the two enzymes supports the theory of the formation of insoluble pectin in the middle lamella as the cause of the occurrence of HTC phenomena in common beans during storage in conditions of high temperature and relative humidity (Jones & Boulter, 1983).
Several studies have been carried out to map QTLs that control cooking time.A random amplified polymorphic DNA (RAPD) marker associated with cooking time was identified using 104 recombinant inbred lines; the identified marker explained 23% of the variation in cooking time (Jacinto-Hernandez et al., 2003).Garcia et al. (2012) mapped six QTLs that govern cooking time on Chromosomes 1 and 9 using 105 polymorphic SSR markers and 140 F 2:4 recombinant inbred lines.The study found QTL CT1.1 on Chromosome 1, which explained 21% of the phenotypic variation to be the most promising.
In a GWAS using freshly harvested seeds, significant SNPs associated with cooking time were identified on Chromosomes 2, 3, and 6 using 206 common bean accessions of Andean origin; the SNPs explained between 4% and 8.7% of the phenotypic variation (Cichy et al., 2015).A recent study conducted by Berry et al.

2. 1 |
Plant material and field multiplication A total of 222 common bean accessions were used in this study and included 169 accessions from the National Gene Bank of Kenya based at Kenya Agricultural and Livestock Research Organisation (KALRO)-Muguga, 38 accessions from KALRO-Embu, 11 landraces collected from farmers' fields, and four commercial varieties (GLP2, GLP24,

of
Agriculture and Technology (JKUAT) trial farm in Kenya.The field trial was conducted as a randomized complete block design with three replicates.The plot size was a single row line of 5 m in length with an inter-row plant spacing of 40 cm and intra-row spacing of 20 cm.The site coordinates are 3 35 0 South and 36 35 0 East at an elevation of 1520 m above sea level and experience a bimodal pattern of rainfall with an annual mean of 856 mm.Wet seasons occur between March and DNA isolation and genotyping were conducted at SEQART Africa laboratories housed in the Biosciences eastern and central Africa (BecA), International Livestock Research Institute (ILRI) Campus, Nairobi.
T A B L E 1 Mean values of cooking time for top ten and bottom five accessions ranked based on cooking time of aged seeds

F
I G U R E 1 Dendrogram showing relationship among 194 common bean genotypes used in this study F I G U R E 2 Population structure of common bean accession, (A) amount of variation accounted for by principal components and (B) threedimensional plot from principal component analysis for 222 common bean accessions and 1988 single nucleotide polymorphisms (SNPs) markers When we explored for potential candidate genes in the Phytozome database, three potential candidate genes were identified near the location of the most significant SNP marker, and eight potential candidate genes were identified around the second most significant SNP marker at around 100 kb of the location of the SNP markers.The nearest genes to the location of the most significant SNP at Phytozome database in Chromosome 10 were Phvul.010G038000 at 5575958 bp, Phvul.010G038100.1 at 5627743 bp, and Phvul.010G038200.1 at location 5,644,933 bp (Table 4).A BLASTn search of the most significant SNP 100096770jFj0-21:G > A-21: G > A sequence (TGCAGTACCAGAAAAACAATCGGTTGTTTTA-CAAAAACAATCGGTTGTTTTACAAAAACAATCGGTTGT) at the NCBI database revealed Sequence ID AC254323.1 and AC254327.1 of P. vulgaris L. with a 90.5% and 89.1% match, respectively.The F I G U R E 3 (A) Distribution of markers by correlation (r) in each of the 11 chromosomes, (B) distribution of markers along the 11 chromosomes, (C) distribution of markers by correlation (r) as a function of genetic distance (kb), (D) linkage disequilibrium (LD,r 2 ) decay plot for pairwise markers as a function of genetic distance (kb), (E) distribution of single nucleotide polymorphisms (SNPs) markers by correlation (r), and (F) distribution of 1988 SNP markers by genetic distance (kb) for the 222 common bean accessions F I G U R E 4 (A) Quantile-quantile plot of estimated Àlog (P) for association analysis of cooking time.(B) Manhattan plot showing significant single nucleotide polymorphisms (SNPs) and their p values for cooking time of aged and soaked common bean accessions functional annotation for transcript Phvul.010G038000 indicated galacturan 1,4-alpha-galacturonidase enzyme, polygalacturonase/ pectinase enzyme for Phvul.010G038100.1,and NADPH-dependent alkenal reductase enzyme for Phvul.010G038200.1.There were no defined functions for the genes identified in the NCBI database.Most of the genes identified in the Phytozome database around the second significant SNP marker SNP 3377419jFj0-24:A > T-24: A > T did not have a defined function; the nearest transcript to the marker was Phvul.010G030800.1 with a functional annotation of asparagine tRNA ligase enzyme, which belongs to Class II family of tRNA synthetases localized in the cytoplasm where it plays a role in protein synthesis.A BLASTn in search of the sequence of the second most significant SNP marker (TGCAGTTCAGGATCTGAAGAAAAC AAATGACCTGGCATCACAATTTGAAGCAAGAGAAAACAGAAAGTT) The cooking time profile in this study agrees with the findings of Kinyanjui et al. (2015) who T A B L E 4 Significant SNPs for cooking time for aged and soaked common bean accessions and their corresponding potential candidate genes in Phytozome and NCBI database For a rapid and effective breeding program, genetic markers are used to identify QTLs for desirable traits.In this study, GWAS was used to identify the genomic regions that control cooking time in common beans.The lack of significant SNP associated with the cooking time of soaked freshly harvested seeds could be attributed to insufficient variation in the cooking time among fresh common bean accessions.Two significant SNP markers were identified to be associated with the cooking time of soaked aged seeds.The study identified two positional potential candidate genes Phvul.010G038000 and enzymes, respectively, close to the position of the most significant SNP marker.The two enzymes are known to be involved in the breakdown of pectin in the plant cell wall.

(
2020) using freshly harvested 146 recombinant inbred lines of common bean identified 10 QTLs on Chromosomes 1, 2, 3, 5, 6, 10, and 11, with the most robust QTLs being in Chromosomes 3, 6, 10, and 11 that appeared in over two different environments.The variations observed in these studies could have resulted from a different number of markers in some studies that affect markers' saturation, pretreatments of seeds before the determination of cooking time such as storage conditions and soaking, and the limitations and advantages of different types of mapping populations used in the studies.5 | CONCLUSIONThis study assessed variation for the cooking time of fresh and aged common bean accessions.In addition, genotyping of the common bean accessions was conducted to identify SNP markers associated with cooking time.There was an increase in cooking time due to storage, which varied among common bean accessions.GWAS analysis revealed two significant SNP markers associated with the cooking time of the aged common bean accessions on Chromosome 10.Consequently, two potential candidate genes colocalized with the most significant SNP marker.The two genes Phvul.010G038000 and Phvul.010G038100encode galacturan 1,4-alpha-galacturonidase and polygalacturonase enzymes, respectively.The two enzymes are involved in the hydrolysis of pectin in the plant cell wall.The research findings of this study support the theory of the formation of insoluble pectin as the cause of the hard-to-cook trait.The characterized common bean accessions and the QTL identified in this study can be utilized in breeding programs to improve the cooking quality of the common bean.
Mean values of cooking time for top ten accessions least affected by storage The DArTseq produced 26,945 SNP markers of which 24,878 were successfully assigned chromosomes; after filtering out SNPs with MAF of <5% and missing data >20%, a final total of 19,188 markers remained, which were subjected to marker-trait association analysis.The average number of SNP markers per chromosome was 1744.4,whichranged from 1387 markers on Chromosome 4 to 2325T A B L E 2Abbreviations: CW, Canadian wonder; CV, coefficient of variation; LSD, least significant difference; min, minutes; RH, red haricot; SC, semi climber.**Significant at 0.01 level.T A B L E 3 Pearson correlation coefficient between cooking time and seven agronomic traits of common bean