MACHETE – algorithm for automated detection of RNA gene fusions

Researchers in Prof. Julia Salzman's laboratory have developed a sensitive, specific algorithm for automated, high-throughput detection of RNA fusions from RNA-Seq data. This technology, called MACHETE (Mismatched Alignment ChimEra Tracking Engine), uses statistical analysis to detect fusion events including splicing between exons separated by more than 100kb or splicing of genes on different chromosomes. It creates a custom database of all exon-exon junctions and prioritizes fusion transcripts for further study by weeding out false positives while retaining true positives. This approach removes false positives from biochemical and bioinformatics artifacts, resulting in unprecedented specificity (no false positives when run on gold standard simulated data) with sensitivity comparable to the best performing published algorithms. Furthermore, the algorithm can analyze short reads and does not require human-guidance for filtering results. MACHETE could potentially pave the way for unbiased de novo discovery of potentially driving and druggable gene fusions in primary tumors with applications in research, diagnostics and drug development.

Stage of Research
The inventors have benchmarked MACHETE's sensitivity and specificity against the best performing published algorithms. In addition, they have used MACHETE to mine public data to discover and validate novel gene fusions in ovarian cancer.