Table 4.
Evaluation of the 3pRule + bootstrap approach for detecting SNP–SNP interactions based on a dataset with 614 SNPs.
| Method1 | Criterion | TPR2 (%) | FPR2 (%) | Total | No. significant pairs1 | ||||
|---|---|---|---|---|---|---|---|---|---|
| C–C (causal pair) | Total null pair | C–N (% null pairs) | C–C-other | N–N | |||||
| 1pRule | Original | 100 | 0.82 | 1538 | 7 | 1531 | 1482 (96.8) | 49 | 0 |
| ≥ 75% runs3 | 100 | 0.24 | 468 | 7 | 461 | 439 (95.2) | 22 | 0 | |
| ≥ 80% runs3 | 85.7 | 0.18 | 346 | 6 | 340 | 325 (95.6) | 15 | 0 | |
| ≥ 90% runs3 | 71.4 | 0.14 | 262 | 5 | 257 | 245 (95.3) | 12 | 0 | |
| 3pRule | Original | 100 | 0.35 | 672 | 7 | 665 | 624 (93.8) | 41 | 0 |
| ≥ 75% runs3 | 100 | 0.04 | 86 | 7 | 79 | 71 (89.9) | 8 | 0 | |
| ≥ 80% runs3 | 85.7 | 0.02 | 48 | 6 | 42 | 37 (88.1) | 5 | 0 | |
| ≥ 90% runs3 | 71.4 | 0.01 | 31 | 5 | 26 | 23 (88.5) | 3 | 0 | |
1Significance rules: 1pRule: p-pair < 2.7 × 10–7; 3pRule: p-pair < 2.7 × 10–7 and p-pair < p-main for SNP1, and p-pair < p-main for SNP2; “C”: a SNP in a causal pair; “N”: a null SNP.
2TPR (true positive rate): proportion of significance out of 7 causal pairs; FPR (false positive rate): proportion of significance null pairs out of all 188,184 null pairs.
3Based on 500 bootstrap runs.