Abstract: To propose a testing method based on Permutation test and Principle Component AnalysisPermutationPCA to explore the performance of PermutationPCA on genotyped data sets under different genetic model assumptions in a casecontrol study. MethodsBased on the control data of the ‘heroininduced positive responses on the first use’ genotyped data, simulated casecontrol genotyped data sets under seven different genetic model assumptions were generated. The PermutationPCA method was used for the simulated data, and also for the real genotyped data of the ‘heroininduced positive responses on the first use′. ResultsThe results of the PermutationPCA method under seven different genetic model assumptions were all significant(Pvalue＜0.05), and the result of the PermutationPCA method on the real genotyped data set was significant (Pvalue＜0.05). ConclusionThe PermutationPCA method is not sensitive to different genetic model assumptions. It is applicable to casecontrol genotyped data in a complex disease gene mapping study.

Key words: Association study, Casecontrol, Genetic model, Principle component analysis