View article

Variables in genetic datasets are often correlated and much larger in number than the sample size, raising the question of how to perform time-efficient multiple testing corrections. We address the problem through a timely efficient algorithm that performs permutation testing in the setting of candidate gene association studies. Permutation analysis represents a statistically sound approach as it yields exact corrections. Compared to existing software, our algorithm is often tens to hundreds times faster. PROC MULTTEST, from the SAS Inc., is the only existing procedure that time-wise performs better than our program. However, in the candidate gene setting, our program grants more flexibility than PROC MULTTEST through several user-defined options. We illustrate the different types of analyses allowed through IGFI data obtained from the Multiethnic Cohort Study. We perform permutation corrections with SNP and …