SNPs were called using Analysis Power Tools 2.10.2.2 (Affymetrix, https://www.affymetrix.com/support/developer/powertools/changelog/index.html ). The resulting genotype.vcfs were scanned with ‘snpflip’ (https://github.com/biocore-ntnu/snpflip ) using the GRCh37 build of the human genome reference sequence maintained by the University of California, Santa Cruz (http://hgdownload.cse.ucsc.edu/goldenPath/hg19/bigZips/hg19.fa.gz ) to identify reversed and ambiguous-stranded SNPs, which were flipped and removed (respectively) using Plink 1.90 (http://pngu.mgh.harvard.edu/purcell/plink/ ) (10 (link)), and the remaining sites were sorted using Plink 2.00a3LM (www.cog-genomics.org/plink/2.0/ ) (11 (link)). This SNP data was then augmented with additional sites imputed by the Michigan Imputation Server (https://imputationserver.sph.umich.edu ) (1000G Phase 3 v5 GRCh37 reference panel, rsqFilter off, Eagle v2.4 phasing, EUR population). SNP positions were translated to GRCh38 coordinates using the ‘LiftoverVcf’ command of Picard 2.23.3 (http://broadinstitute.github.io/picard/ ). Finally, Vcftools 0.1.13 (12 (link)) was used to exclude non-exonic SNPs and SNPs with minor allele frequency < 0.05.
Free full text: Click here
