Issues with sample quality can be easily identified by the proportion of SNPs that fail to have their genotype called, and samples with a low call rate (usually more than 7% SNP genotypes missing) should be disregarded from the analysis. There are other sources of errors that may impact the analysis and are less evident. Experience shows that inefficient sample tracking—for example not using bar code reading systems to track samples—or sample swaps is unfortunately more frequent than lab technicians would like to admit [62]. There are at least two analyses that can be easily conducted to point to possible errors. The first analysis examines the agreement between gender specification in the phenotypic data and the gender inferred by heterozygosity of the SNPs on chromosome X. The second analysis consists of scoring the genetic similarity between every pair of subjects in the study using estimates of the alleles that are shared identity by descent (IBD). Two alleles are shared IBD when they are identical copies of the same ancestral allele. Mendel’s laws of inheritance can be used to estimate the
