Skip to main content

View full-text article in PMC

. 2016 Sep 6;8(1):92. doi: 10.1186/s13148-016-0256-8

Table 1.

Estimated cohort sizes for 80 % power at individual loci and genome wide in twin and case-control EWAS designs (from [93], under CC BY 4.0 licence)

Difference (%)	Twin				Case-control
	P < 0.05		P < 1 × 10⁻⁶		P < 0.05		P < 1 × 10⁻⁶
	t test^a	Wilcox^b	t test^a	Wilcox^b	t test^c	Wilcox^d	t test^c	Wilcox^d
7	30	30	178	178	37	37	211	211
8	25	25	145	149	30	30	169	169
9	20	20	117	117	24	24	137	137
10	17	18	98	102	20	21	112	110
11	15	15	81	83	17	18	96	95
12	13	13	71	71	15	16	80	80
13	11	12	63	69	13	13	70	70
14	10	11	55	62	11	13	61	63
15	9	10	50	57	10	11	54	57

^a t test, paired t test

^bWilcox, Wilcoxon signed-rank test

^c t test, two-sample t test

^dWilcox, Wilcoxon rank-sum test