Table 1 Optimal SNP subsets using LDA or sIB for predicting psoriasis with average accuracy and 95% confidence interval estimated from Bootstrap re-sampling
Subsets | Components (dbSNP_rs on chromosome) | CV HMSS (Bootstrap mean and 95% CI) | Total CV accuracy (Bootstrap mean and 95% CI) | Test HMSS (Bootstrap mean and 95% CI) | Total test accuracy (Bootstrap mean and 95% CI) |
|---|---|---|---|---|---|
LDA | Â | Â | Â | Â | Â |
1 SNP* | rs10905106 on 10 | 0.498(0.498, 0.475-0.518) | 0.495(0.496, 0.474-0.516) | 0.544(0.494, 0.469-0.519) | 0.553(0.508, 0.486-0.532) |
2 SNPs* | rs10958357 on 8 rs7973936 on 12 | 0.486(0.499, 0.480-0.520) | 0.495(0.500, 0.481-0.521) | 0.556(0.498, 0.471-0.525) | 0.565(0.512, 0.491-0.535) |
1 SNP∆ | rs4375421 on 11 | 0.540(0.497, 0.474-0.519) | 0.540(0.498, 0.476-0.518) | 0.492(0.500, 0.476-0.529) | 0.491(0.499, 0.474-0.529) |
2 SNPs∆ | rs950753 on 3 rs7058025 on X | 0.570(0.493, 0.468-0.514) | 0.575(0.508, 0.486-0.530) | 0.463(0.476, 0.451-0.502) | 0.459(0.473, 0.450-0.498) |
FS | 38 SNPs | 0.604(0.500, 0.478-0.520) | 0.622(0.503, 0.482-0.523) | 0.520(0.496, 0.472-0.524) | 0.514(0.491, 0.466-0.518) |
SFFS | 32 SNPs | 0.622(0.497, 0.475-0.520) | 0.622(0.502, 0.479-0.525) | 0.512(0.498, 0.472-0.523) | 0.509(0.498, 0.473-0.522) |
sIB | Â | Â | Â | Â | Â |
1 SNP* | rs12191877 on 6 | 0.611(0.605, 0.563-0.630) | 0.611(0.608, 0.580-0.631) | 0.668(0.668, 0.641-0.694) | 0.699(0.698, 0.676-0.720) |
2 SNPs* | rs12191877 on 6 rs4953658 on 2 | 0.557(0.444, 0.014-0.633) | 0.574(0.550, 0.426-0.633) | 0.674(0.674, 0.650-0.698) | 0.685(0.684, 0.662-0.707) |
FS | rs12191877 on 6 | 0.611(0.605, 0.563-0.630) | 0.611(0.608, 0.580-0.631) | 0.668(0.668, 0.641-0.694) | 0.699(0.698, 0.676-0.720) |
SFFS | rs2844627 on 6 rs7773175 on 6 | 0.619(0.617, 0.576-0.641) | 0.616(0.615, 0.585-0.638) | 0.659(0.658, 0.633-0.683) | 0.677(0.676, 0.655-0.699) |