Power for complex trait genetic association

One of the key issues facing researchers who want to map genes for complex traits is appropriate methodology for statistical power calculations. Most classic methods assume that parameters for the genetic model are known, which is rarely the case for complex traits. Furthermore, few if any methods use empirical data from genes of interest. We present a statistically valid method for performing such power calculations using empirical data and apply it to a candidate gene example for schizophrenia. We also document several advantages of our method, most notably the computation speed with which our power calculations may be performed.