Gene association analysis via the method of local polynomial regression

Conference: BIBE 2019 - The Third International Conference on Biological Information and Biomedical Engineering
06/20/2019 - 06/22/2019 at Hangzhou, China

Proceedings: BIBE 2019

Pages: 5Language: englishTyp: PDF

Personal VDE Members are entitled to a 10% discount on this title

He, Jinli; Guo, Yixing; Li, Qingqing; Zhao, Xiaohan (Department of Statistic, School of Mathematical Sciences, Heilongjiang University and Heilongjiang Provincial Key Laboratory of the Theory and Computation of Complex Systems, Harbin, China)

Population stratification is an significant issue in both rare variant and common variant association analysis. Failure to correctly account for the bias caused by population stratification will lead to spurious association in structured population. In the present paper, we proposed a PC-based local polynomial regression (PC-lpr) method to correct population stratification for both rare variant and common variant association test. Furthermore, we compare the PC-lpr method with the methods Uncorrected, GC, PC-linear and PC-nonp in simulation studies. Simulation results suggest that the PC-lpr method outperforms other methods in the adjustment of population stratification. We also find that the PC-lpr method has higher power in the most situations.