For this contig, Hapler called 2,845 SNPs (using the "Simple Strict" SNP caller, see Methods) and reconstructed a consensus with 146 estimated crossovers in 12 minutes 38 seconds.
The genotype calls are likely sufficiently reliable since comparisons between r2 values generated from this SNP calling method and from the stricter SNP calling method described below under "Vitis9KSNP array" were highly correlated (r2 = 0.95, p<1×10−15).
Strict criterion SNPs occurred in 2,067 putative E. propertius coding regions, representing 1.52 Mbp of sequence.
For E. propertius, strict criterion SNPs were found in 6,298 contigs, comprising 6.11 Mbp of sequence.
If min_len_downstream and min_len_upstream are greater than zero, then this is a strict definition of SNPs.
After strict filtering of SNPs on the basis of read depth and minor allele frequency (MAF > 0.05), a final set of 8,397 SNPs (InDels were not used) was obtained and used for association analysis.
Due to the strict three-step SNP-based, window-based, and combined analysis, when we did the literature search we were able to find direct or indirect support from previous studies for the majority of our predicted CNVs (see Table 2 for detailed validation information), which indicated the meaningfulness of our CNV predictions.
However, we applied a targeted approach, and if we had used a more strict selection of putative SNPs this would, in our case, decreased the total number of putative SNPs rather than led to selection of other putative SNPs in the target genes that were more likely to pass the genotype validation.
This GWAS was based on 502,702 SNPs after strict QC.
The transferability in the strict sense, namely those SNPs that maintained the heterozygous state, was analysed based on the four grapevine sub-groups (24 V. vinifera cultivars, 10 V. vinifera subsp. sylvestris accessions, 35 non- vinifera Vitis accessions, and 2 M. rotundifolia accessions).
