*4.4. Analysis of Genetic Structure*

The STRUCTURE software version 2.3.4 [79] was employed to define 132 accessions into clusters consisting of genotypes by using co-dominant genotypic data. The admixture model was used to estimate a mixed group by using correlated allele frequencies between populations as described by Falush et al. [80]. The optimum number of subpopulations was calculated based on the recommendation of Evanno et al. [81] by defining the values for *K* = 2 to *K* =10 with a burn length of 10,000 and a run length of 100,000 each in 10 runs. The results were uploaded in a Zip file to the STRUCTURE harvester software for finding the ∆*K* [82]. Grouping and subgrouping of accessions

was done if the probability of membership was more than 70% [83]. The accessions with membership <70% were placed into the mixed subgroup.
