Table 1.
Method comparison using simulated data
| Method | Mean number of detected QTL | Mean number of false positives | Ratio |
|---|---|---|---|
| Sliding-5 | 6.33 | 6.39 | 0.990610329 |
| Sliding-10 | 13.29 | 4.16 | 3.194711538 |
| Sliding-25 | 17.6 | 75.1 | 0.234354194 |
| Sliding-50 | 18.38 | 232.58 | 0.079026572 |
| Sliding-100 | 18.23 | 488.19 | 0.037342018 |
| Sliding-250 | 18.41 | 2082.54 | 0.008840166 |
| Sliding-500 | 19.51 | 9065.05 | 0.002152222 |
| Distinct-5 | 7.42 | 0.12 | 61.83333333 |
| Distinct-10 | 12.19 | 0.99 | 12.31313131 |
| Distinct-25 | 16.75 | 7.67 | 2.183833116 |
| Distinct-50 | 18.01 | 11.27 | 1.598047915 |
| Distinct-100 | 17.73 | 9.52 | 1.862394958 |
| Distinct-250 | 19.49 | 31.53 | 0.618141453 |
| Distinct-500 | 18.34 | 28.47 | 0.644186863 |
| Spline Windows | 15.98 | 3.4 | 4.7 |
Results from applying an assortment of window-methods applied to 100 simulated selection experiments involving 30 QTL, 30 generations of selection, and pooled sequencing at 1 000 000 markers to estimate allele frequencies. The mean number of QTL (out of 30) detected over the 100 simulations, mean number of false positives, and ratio of detections to false positives across simulations is provided for each of the methods evaluated. Sliding- and Distinct- refer to sliding and distinct window methods with windows of the specified size, and Spline Windows refers to the method described here and employed in GenWin, where window size is not restricted a priori.