Table 2.
Fusion discoveries in the spike-in dataset. The results for TopHat-Fusion, SOAPfuse and JAFFA for sample SRR1659964 are collected from a recent study [9]
| One sample(SRR1659964) | 20 samples | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| TP | Total | Precision | Recall | F1 | P-value | TP | Other | P-value | |
| FusionMap | 7 | 26 | 0.27 | 0.78 | 0.40 | 0.78 | 142/180 | 283 | 0.04 |
| TRUP | 4 | 9 | 0.44 | 0.44 | 0.44 | 1 | 80/180 | 63 | 0.15 |
| JAFFA | 5 | 13 | 0.39 | 0.56 | 0.46 | 1 | 138/180 | 114 | 0.12 |
| SOAPfuse | 4 | 13 | 0.31 | 0.44 | 0.36 | 1 | NA | NA | NA |
| TopHat-Fusion | 1 | 6 | 0.17 | 0.11 | 0.13 | 0.66 | 133/180 | 925 | 1.6e-22 |
| FuSeq | 9 | 25 | 0.36 | 1 | 0.53 | - | 179/180 | 228 | - |
TP= true positive fusion genes, Other= unvalidated fusion genes, Total= total discovered fusion-gene candidates, P-value = two-sided p-value of Fisher’s exact test of the difference in precision between FuSeq vs each of the other methods