. Author manuscript; available in PMC: 2018 Oct 23.

Published in final edited form as: Nat Genet. 2018 Apr 23;50(5):693–698. doi: 10.1038/s41588-018-0099-7

Table 1:

Power to detect horizontal pleiotropy in Mendelian randomization for different methods.

Causal effect	Percent	Pleiotropy	MR-PRESSO global test	Q (modified) test	Q’ (modified) test
0	2	balanced	25.34	25.40	22.04
0	2	positive	25.01	25.01	22.00
0	4	balanced	51.79	51.96	47.80
0	4	positive	50.88	51.32	47.34
0	10	balanced	95.53	95.58	94.47
0	10	positive	94.27	94.26	92.85
0.1	2	balanced	24.10	24.41	21.58
0.1	2	positive	23.60	23.89	20.96
0.1	4	balanced	51.17	51.49	47.51
0.1	4	positive	50.67	50.59	46.69
0.1	10	balanced	95.56	95.56	94.31
0.1	10	positive	93.73	93.77	92.08
0.2	2	balanced	22.42	22.70	19.74
0.2	2	positive	22.95	22.92	19.97
0.2	4	balanced	48.37	48.29	44.45
0.2	4	positive	46.89	46.82	43.18
0.2	10	balanced	94.08	94.11	92.78
0.2	10	positive	91.85	91.91	90.24
0.5	2	balanced	16.72	16.70	14.90
0.5	2	positive	16.76	16.55	15.15
0.5	4	balanced	33.71	33.79	31.00
0.5	4	positive	32.52	32.66	29.93
0.5	10	balanced	81.15	81.22	79.07
0.5	10	positive	76.99	77.09	74.75

A total of 50 variants was simulated in each case. The InSIDE (Instrument Strength Independent of Direct Effect) condition is satisfied in all reported scenarios. MR-PRESSO: Mendelian Randomization Pleiotropy RESidual Sum and Outlier.