Currently the laboratory methods used for the 10k Affymetrix assay include multiple manual steps involving the movement of DNA and reagents via a single-channel or multi-channel pipettes. As a result, one sample swap occurred in the CIDR lab. This sample problem was detected when checking for Mendelian inconsistencies, sex, and replicate errors. Identity by state sharing was calculated within families and across all samples in the dataset as an additional method to screen for problematic samples. Suspect samples were re-genotyped. Confirmation that the problems were resolved was achieved by checking lab-to-lab and within-lab replicates as well as confirming genotype concordance for 26 SNPs in common between the Affymetrix and Illumina datasets.
