Table 3. Power of different methods in a disease model in which 20% of causal variants are protective.

Error rate	Coverage per person	C-alpha	Freq-weight	VT	RWAS	LRT_G	VST	LRT
0	4	0.054	0.417	0.653	0.471	0.646	0.456	0.956
	10	0.069	0.46	0.712	0.472	0.689	0.506	0.987
	20	0.07	0.473	0.713	0.478	0.699	0.515	0.99
0.01	4	0.053	0.14	0.164	0.17	0.191	0.222	0.685
	10	0.066	0.304	0.468	0.376	0.495	0.425	0.955
	20	0.06	0.458	0.673	0.465	0.67	0.491	0.976

PAR was 0.02 with c_i = 0.1 Tested methods were C-alpha (Neale et al. 2011), Freq-weight similar to the Madsen–Browning method (Madsen and Browning 2009), variable threshold (VT) (Price et al. 2010), RWAS (Sul et al. 2011b), LRT_G (Sul et al. 2011a), and our proposed methods (VST and LRT).