Detecting rare functional variants using a wavelet-based test on quantitative and qualitative traits

BMC Proceedings

Table 1 Powers and type I error rates

Window	Q1	Q2	Q4	Affected
16 SNPs
Power	FLT1 (0.82) KDR (0.17) ARNT (0.015) VEGFC (0.005)	BCHE (0.02) SIRT1 (0.005) VNN3 (0.035) LPL (0.005) VLDLR (0.01) SREBF1 (0.02)		HSP90AA1 (0.005) PRKCA (0.005)
Type I error	0.085	0.121	0.134	0.099
8 SNPs
Power	FLT1 (0.68) KDR (0.16)	BCHE (0.01) SIRT1 (0.005)		HSP90AA1 (0.005) PRKCA (0.005)
Type I error	0.035	0.075	0.11	0.05
4 SNPs
Power	FLT1 (1.0) KDR (0.15)	RARB (0.005) VNN1 (0.025) VNN3 (0.01) SREBF1 (0.005)		SOS2 (0.005)
Type I error	0.247	0.162	0.178	0.150
Genes
Power	FLT1 (0.68) KDR (0.26)	GCKR (0.01) BCHE (0.01) VNN1 (0.02) LPL (0.005) VLDLR (0.01) SIRT1 (0.005) SREBF1 (0.005)		HSP90AA1 (0.005) PRKCA (0.005)
Type I error	0.06	0.21	0.27	0.155
4 less rare SNPs, no adjustments for other factors
Power	FLT1 (0.575) ARNT (0.015)	VNN1 (0.005) VNN3 (0.005) LPL (0.015)
Type I error	0.050	0.136	0.085	0.105

Using a gene as the window does not make any adjustment for population stratification and environmental factors. Values in parentheses are the powers.

ISSN: 1753-6561