Abstract:
|
Traditional statistical methods analyzing imaging genetics data often have low statistical power. We previously proposed a method called Functional Mixed Effect Model (FMEM) with a likelihood ratio test, which provides better power to identify the genetic effect to the imaging responses while incorporating the linkage disequilibrium (LD) in genomic information, and spatial correlation in imaging responses. However, due to high dimensional feature in imaging data, the method is computationally intensive. Thus we further proposed fast algorithm with a weighted score test statistic, which pertains the correlation in both genomic variation and imaging responses with better statistical power. Our method is evaluated by simulation studies and the result shows it outperforms voxel-wise based approach by greater statistical power with great type I error control with efficient computational time. To our up-to-date knowledge, our method is the first one to directly weight the original test statistic instead of weighting likelihood or estimators.
|
ASA Meetings Department
732 North Washington Street, Alexandria, VA 22314
(703) 684-1221 • meetings@amstat.org
Copyright © American Statistical Association.