Table 2 Improved precision in fusion detection, utilizing Seraseq controls, achieved by EnFusion. Data shown is from undiluted Seraseq v3 RNA-Seq, experiments performed in duplicate, averages are shown. Individual algorithms are listed by precision, in descending order. Seraseq fusions identified (true positive) are out of a possible 14 fusions
From: Discovery of clinically relevant fusions in pediatric cancer
Algorithm | Total fusions identified | Seraseq fusions identified | Sensitivity | Precision |
|---|---|---|---|---|
Arriba | 23.5 | 13 | 92.9% | 55.3% |
MapSplice | 22 | 12 | 85.7% | 54.6% |
STAR-Fusion | 32 | 14 | 100.0% | 43.6% |
FusionMap | 30 | 12.5 | 89.3% | 41.7% |
FusionCatcher | 299.5 | 13 | 92.9% | 4.3% |
JAFFA | 470.5 | 12.5 | 89.3% | 2.7% |
CICERO | 1323 | 14 | 100.0% | 1.1% |
EnFusion 2 callers | 40 | 14 | 100.0% | 35.0% |
EnFusion 2 callers + filter | 15.5 | 12 | 85.7% | 77.4% |
EnFusion 2 callers + filter + known fusion list | 17.5 | 14 | 100.0% | 80.0% |
EnFusion 3 callers | 15.5 | 14 | 100.0% | 90.3% |
EnFusion 3 callers + filter | 12 | 12 | 85.7% | 100.0% |
EnFusion 3 callers + filter + known fusion list | 14 | 14 | 100.0% | 100.0% |