A Bayesian model for detection of high-order interactions among genetic variants in genome-wide association studies

Table 1 Simulation results on epistasis models 1–4

Model	Sample	LD	MAF	BHIT	BHIT	PLINK	PLINK	PLINK	BOOST	EDCF
Model	number	LD	MAF	(One)	(Triple)	PLINK	(Fast)	(Q)	BOOST	EDCF
1	2000	1	0.5	1.00	1.00	0.98	0.34	1.00	1.00	1.00
			0.1	0.94	1.00	0	0	0.90	0	0
		0.7	0.5	0.90	1.00	0.22	0.04	0.68	0.84	0.66
			0.1	0.92	1.00	0	0	0.48	0	0
	4000	1	0.5	1.00	1.00	1.00	0.84	1.00	1.00	0.80
			0.1	0.94	1.00	0	0	1.00	0.22	0.90
		0.7	0.5	0.86	1.00	0.68	0.18	1.00	1.00	0.38
			0.1	0.88	1.00	0.04	0.04	0.96	0.12	0.82
2	2000	1	0.5	1.00	1.00	0.72	0.38	0.98	1.00	1.00
			0.1	0.56	0.74	0	0	0	0	0
		0.7	0.5	0.88	1.00	0.22	0.08	0.48	0.42	0.80
			0.1	0.20	0.22	0	0	0	0	0
	4000	1	0.5	0.98	1.00	1.00	0.88	1.00	1.00	1.00
			0.1	0.56	0.78	0	0	0	0.16	0.48
		0.7	0.5	0.96	1.00	0.64	0.32	0.92	0.98	0.86
			0.1	0.20	0.24	0.06	0.02	0	0.14	0.28
3	2000	1	0.5	0.94	1.00	0.38	0.10	0.82	0.98	1.00
			0.1	0.30	0.34	0	0	0	0	0
		0.7	0.5	0.94	1.00	0.04	0.32	0.32	0.16	0.74
			0.1	0.02	0.04	0	0	0	0	0
	4000	1	0.5	1.00	1.00	0.96	0.62	1.00	1.00	1.00
			0.1	0.24	0.32	0	0	0	0.02	0.42
		0.7	0.5	0.90	1.00	0.28	0.12	0.76	0.60	0.88
			0.1	0.12	0.16	0.04	0	0	0	0.28
4	5000	1	0.5	0.96	1.00	0	0	0	0.04	0.14
			0.1	0.08	0.16	0	0	0	0	0.08
		0.7	0.5	1.00	1.00	0	0	0	0	0.08
			0.1	0.02	0.04	0	0	0	0	0
	10,000	1	0.5	1.00	1.00	0	0	0	0.32	0.42
			0.1	0.66	0.94	0	0	0	0	0.02
		0.7	0.5	0.90	1.00	0	0	0	0.02	0.58
			0.1	0.58	0.82	0	0	0	0	0

Simulation results of statistical power on each of four defined models in each simulation setting. Each simulation contains 50 simulation datasets generated by each model. The sample sizes of Models 1–3 are set as 2000 and 4000, and the sample size of Model 4 is set as 5000 and 10,000. The LD between variations and ground-truth loci is set as 1 or 0.7. The MAF is set as 0.7 or 0.1. Bold shows highest statistical power

ISSN: 1471-2164