Gene-based multiple trait analysis for exome sequencing data

BMC Proceedings

Table 1 Power to detect the causal genes on Q1 at the 0.05 significance level

Gene	Q1	Q2	Q4	D	Q1+D	Q1+Q2 +Q4+D	OCMT	Subset with the highest power (power)
ARNT	0.185 0.145	0.050 0.050	0.055 0.040	0.070 0.065	0.180 0.165	0.140 0.115	0.160 0.135	Q1+D (0.180) Q1+D (0.165)
ELAVL4	0.520 0.045	0.050 0.035	0.060 0.045	0.030 0.045	0.465 0.040	0.390 0.040	0.380 0.045	Q1+D (0.465) Q1+Q4 (0.055)
FLT1	1.000 0.995	0.075 0.075	0.055 0.040	0.700 0.475	1.000 0.995	1.000 0.985	1.000 0.980	Q1+D (1.000) Q1+Q2 (1.000)
FLT4	0.865 0.030	0.055 0.040	0.060 0.060	0.290 0.085	0.765 0.065	0.600 0.045	0.655 0.045	Q1+D (0.765) Q2+D (0.085)
HIF1A	0.565 0.025	0.220 0.080	0.040 0.020	0.280 0.095	0.530 0.055	0.450 0.060	0.465 0.065	Q1+D (0.530) Q2+D (0.090)
HIF3A	0.030 0.090	0.055 0.040	0.065 0.060	0.030 0.050	0.010 0.080	0.035 0.065	0.045 0.060	Q2+Q4 (0.060) Q1+Q4 (0.085)
KDR	1.000 0.995	0.115 0.020	0.055 0.025	0.680 0.425	1.000 0.990	1.000 0.985	1.000 0.985	Q1+D (1.000) Q1+D (0.990)
VEGFA	0.525 0.305	0.065 0.040	0.055 0.040	0.115 0.090	0.460 0.230	0.320 0.135	0.335 0.135	Q1+D (0.460) Q1+D (0.230)
VEGFC	0.785 0.720	0.050 0.045	0.065 0.050	0.325 0.300	0.790 0.665	0.600 0.490	0.625 0.530	Q1+D (0.790) Q1+D (0.665)

We report the powers without (upper) and with (lower) the adjustment of stratification for the single-trait analyses (Q1, Q2, Q4, D), the CMT method for Q1 and D (Q1+D), the CMT method for all the four traits (Q1+Q2+Q4+D), and the OCMT method (OCMT). The last column presents the subset with the highest power among all the subsets with at least two traits and its power (in parentheses).

ISSN: 1753-6561