We induced population stratification in a hypothetical exome sequencing study involving real sequencing data and simulated phenotype data using 184 individuals from the HIV and 254 individuals from Schizophrenia (SCZ) exome sequencing studies. We observed that there were exome-wide differences in allele frequencies between the populations, which we quantified by estimating the FST between HIV and SCZ samples using exome sequencing data95. FST was estimated using the EIGENSOFT software. Using variants with minor allele frequencies at least 5%, we observed an FST value of 0.003, which is consistent with the different European ancestries of the HIV (European-American) and SCZ (Swedish) samples and with previous estimates of genetics distances between European populations96. We considered the possibility that the observed differences between HIV and SCZ samples could be due to differential bias resulting from differences in sample collection, sequencing, or data processing97, but view this as unlikely because we applied identical data processing and QC procedures to both sample sets and because QC metrics revealed no systematic differences between the sample sets.
