Skip to main content
. 2021 Oct 7;22:484. doi: 10.1186/s12859-021-04356-5

Table 1.

Power of the different methods depending on the number of components in the ‘Random LD Simulations’ (dilution effect)

Size Model Method Significance 1 2 3 4 5 6–10 11–15 16–20 > 20
1000 MD WS 1×10-5 0.4 0.5 0.8 1.0 1 0.7 0.6 0.7 0.7
1000 RD WS 1×10-5 0.8 0.1 0.0 0.3 0.2 0.5 0.9 0.2 0.4
1000 NA SKAT 1×10-5 0.0 0.0 0.0 0.0 0.0 0.1 0.2 0.8 1.8
1000 NA GWAS 5×10-8 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
5000 MD WS 1×10-5 9.2 11.1 9.1 6.1 4.6 3.5 2.5 2.3 2.3
5000 RD WS 1×10-5 9.6 9.3 7.3 5.4 5.1 2.6 1.4 1.7 1.6
5000 NA SKAT 1×10-5 0.0 0.2 0.57 0.8 2.8 7.14 26.2 54.5 82.0
5000 NA GWAS 5×10-8 45.5 25.8 16 3.1 2.7 2.8 1.1 0.6 0.3
10,000 MD WS 1×10-5 48.1 46.5 44.4 45.1 44.1 31.0 24.6 21.2 21.2
10,000 RD WS 1×10-5 49.1 45.2 46.3 47.2 39.2 27.2 15.6 17.2 15.9
10,000 NA SKAT 1×10-5 0.2 1.1 1.3 4.4 7.3 32.0 71.6 92.8 99.2
10,000 NA GWAS 5×10-8 100 81.8 75.0 68.8 30.6 28.1 15.5 9.6 3.3