For all three samples, the distributions of the WGS genotyping metrics were analyzed and compared within the groups formed based on the concordance, variation class and genotype zygosity criteria (data not shown). While differences in mean values were statistically significant (Welch’s t-test, 0.05 p-value threshold) in several groups, distributions themselves exhibited low intergroup divergence. Moreover, for all group comparisons, observed differences were quite small, and their patterns were not uniform, with several groups being insufficiently large, potentially obscuring inter-group dissimilarity. Thus, discordance in genotyping could not be explained by the variance in these quality metrics. When quality metrics were analyzed for the BeadChip calls, some differences in parameters distributions were found. The distinction was remarkable only for the SNV group, as it included the significant amount of all genotyping results and thus could more precisely represent any actual deviations. Inspection of R, Theta and GC scores for both concordant and discordant variants revealed a pattern of discordant variants located close to the borders of the variant clusters (Fig. 5). Importantly, similar clusters are used in the Illumina GenomeStudio software to
