Results Identifying regions of interest by association testing Under a mixed linear model (MLM) from Genome-wide Efficient Mixed Model Association (GEMMA) software with a relationship matrix, sex, and age as covariates (called Model 1), no regions reached genome-wide significance, although two regions on ECA 1 (ECA 1: 171.8 Mb, P = 4.81 × 10 −7 and ECA 1: 2.5 Mb, P = 5.74 × 10 −7 ) approached significance, based on multiple testing correction for the number of independent tests (α mod.bonf . = 2.05 × 10 −7 ) (Fig. 2 A; Table 1 ).
1
—
—