Genotyping technology and allele calling algorithms continue to improve and quality-improvement strategies continue to ensure that only reliable, rigorously scrutinized markers and samples are used for analysis. Reconciling genetic data with clinical and self-reported data (e.g., sex or familial relationships) can potentially identify sample identity problems caused by sample handling mishaps. Batch effects, population stratification, and sample relatedness can confound genetic association analyses and can lead to excessive type I and type II errors. Here we discuss methods that can be used to detect and account for various data quality issues to better ensure the integrity of the primary GWAS as well as its downstream applications.
