FusionScan

FusionScan detects fusion genes from RNA-Seq data to identify fusion transcripts relevant to oncogenic mechanisms in cancer research.

Key Features:

• High Precision and Recall: Achieves a reported precision of 60% and recall of 79%.
• Optimized Detection Strategy: Focuses on split reads composed of intact exons at fusion boundaries to enhance identification of genuine fusion transcripts.
• Advanced Mapping and Filtering: Implements mapping and filtering strategies to eliminate false positives while preserving true positive detections.
• Performance Validation: Validated on cell line datasets NCI-H660, K562, and MCF-7 with known fusion cases and demonstrated superior performance compared to existing programs.
• Robustness Across Conditions: Simulation tests show maintained high sensitivity and specificity across different sequencing depths and read lengths.
• Efficient Computation Time: Computation time is reported to be comparable to other leading tools.

Scientific Applications:

• Cancer research: Identification of fusion genes to inform understanding of oncogenic mechanisms and potential therapeutic targets.
• Experimental validation: Prioritization of high-confidence fusion transcript candidates for downstream experimental confirmation.

Methodology:

Searches for split reads indicating fusion boundaries and uses 269 known fusion cases as a reference.