The 385365 (0.6%) SNPs and 491247 (45.32%) InDels were found novel when compared against dbSNP build 148. The proportion of novel SNPs is lower than reported in previous cattle resequencing studies, such as in, Fleckvieh, Holstein, Black Angus, Kuchinoshima Ushi, Boran, Ogaden, Kenana, Ankole, N’Dama and Chikso breeds (Choi et al., 2013; Grant, Arantes, Liao, ; Stothard, 2011; Kawahara-Miki et al., 2011; Kim et al.
, 2017; Stothard et al., 2011). The lower novel SNPs proportion may be generally accounted for by the recent SNP depositions from the previous large number of bovine resequencing studies, and the use of stringent variant filtering standards to minimize false positive variant calls. Compare to SNPs, a higher proportion of novel InDels could in part be because a large number of next-generation re-sequencing experiments in cattle have focused on reporting SNPs rather than InDels. Only 1.
6% of the total variation discovered in this study is attributed to InDels. However, the InDels involve nearly 3.8% of the entire variant bases, implying that InDels may be a significant source of both genomic and phenotypic diversity.Among the total SNPs and InDels, the homozygous and heterozygous ratios were 1:1.3 (32890916 verses 34412553 SNPs) and 1:1.27 (479060 versus 604782InDels), respectively.
To evaluate the quality of identified SNPs, the transition (TS) versus transversion (TV) ratio of all detected SNPs was measured and on an average, we found 2.3 TS versus 1 TV (TS: TV 2.3:1) across the entire resequenced genome of all samples under study.
The Ts: Tv ratio value is analogous to the ratios documented in other related studies for cattle and human. (Abecasis et al., 2012; Choi et al., 2013; Daetwyler et al., 2014a). Among 1083842 InDels, 573528 (52.9%) are deletions. The lengths of InDels ranges from -26 (deletion) to 20 (insertion).
However, most of the determined InDels are short: approximately 80.48% of InDels are less than and equal to 3bp (Fig2. a) which is comparable to the previous report (Choi et al., 2013; Kawahara-Miki et al.
, 2011). The minimum depth for InDels was set to 6 supporting reads, resulting in nearly 98.3% of InDels with at least 10 assisting reads in the final call set. In addition, InDels with very high read depth (mean + 3*StdDeviation) were also excluded to minimize the PCR artifacts. These results collectively show that the sequencing data supports the identified InDels.