Table 5. Prediction performance of genetic risk score and family history for glioma risk.
| Datasets | No. of subjects | AUC(95%CI)a | H-L testb P value |
|---|---|---|---|
| Dataset2 | |||
| fmc | 1453 | 0.535(0.515-0.554) | 1.000 |
| cGRS | 1643 | 0.607(0.581-0.644) | 0.386 |
| wGRS1 | 1643 | 0.610(0.583-0.638) | 0.111 |
| wGRS2 | 1643 | 0.611(0.584-0.639) | 0.049 |
| PRFLR(SNPs) | 1453 | 0.615(0.586-0.644) | 0.051 |
| cGRS+fmc | 1453 | 0.620(0.591-0.649) | 0.816 |
| wGRS1+fmc | 1453 | 0.623(0.595-0.652) | 0.334 |
| wGRS2+fmc | 1453 | 0.621(0.592-0.650) | 0.117 |
| PRFLR(SNPs+fmc) | 1453 | 0.625(0.596-0.653) | 0.250 |
| Dataset3 | |||
| fmc | 1718 | 0.526(0.508-0.543) | 1.000 |
| cGRS | 1921 | 0.605(0.580-0.629) | 0.997 |
| wGRS1 | 1921 | 0.607(0.582-0.632) | 0.880 |
| wGRS2 | 1921 | 0.608(0.583-0.633) | 0.113 |
| PRFLR (SNPs) | 1718 | 0.635(0.610-0.660) | 0.927 |
| cGRS+fmc | 1718 | 0.611(0.585-0.637) | 0.743 |
| wGRS1+fmc | 1718 | 0.611(0.585-0.638) | 0.154 |
| wGRS2+fmc | 1718 | 0.609(0.583-0.636) | 0.016 |
| PRFLR(SNPs+fmc) | 1718 | 0.646(0.619-0.672) | 0.393 |
AUC: the area under operating characteristic curves; fmc: family history of caner; cGRS: count genetic risk score; wGRS: weighed genetic risk score; PRFLR: predicted risks from logistic regression analysis; a, 2000 bootstrap replicates; b, Hosmer-Lemeshow “goodness-of-fit” test for model calibration.