|
Abstract:
|
Genetic association studies can be confounded by undetected population structure. Such confounding can lead to spurious associations and failure to detect causal associations. Genomic Control (GC) and family-based methods have been proposed to deal with population structure. They are based on two models for population structure: Isolates and Inbreeding. I will show that confounding by population structure can occur under the Isolates Model but not the Inbreeding Model. Thus, GC methods adjust for variance distortion, but not confounding. Conversely, family-based association studies control for confounding, but not variance distortion.
|
