Version 0.5.1

The main focus on this update is to reduce the number of false positive calls by adjusting default parameters, and introducing new filters. Additional improvements to plotting. The plotting script was not included in tag 0.5.0.

Feature: Addition of minimum fusion fraction filter (--min-fusion-fraction). Increased specificity at a small cost to sensitivity.
Feature: Addition of lncRNA filtering (--gtf-transcript-allow-lncRNA). Increased specificity at a small cost to sensitivity.
Feature: Increasing min map quality to 10
Feature: Increasing min identity to 93%
Feature: Adding transcript count for fusion support (XC), and total count (TC) to output.
Bug fix: Fix fofn hang
Bug fix: Ensure unique fusion IDs

Version 0.4.1

• Bug fix: consistent ordering between gene ids and gene names when dealing with ambiguity.
• Bug fix: rely on breakpoint ordering in visualization script.
• Improvement: gene names are assigned automatically in visualization script.
• Improvement: visualization script can now take SAM files as input.
• Improvement: visualization script can now take a breakpoint number (these aren't always unique, so it's best to still give a single entry).

Version 0.4.0

• Improvement: Improved ambiguous gene resolution.
• Improvement: read lzma-compressed annotations in text and binary format.
• Improvement: Support multiple bam/fofn (File-Of-FileNames) inputs.
• Improvement: reduced false-positive fusions.
• Improvement: Additional metadata for putative fusion candidates.
• Improvement: Updated cached binary format. WARNING: this is a breaking change. Old binary annotation files will need to be re-generated.
• Alteration: reduced chaining distance for clustering breakpoints.
• Alteration: emit sense-antisense events by default.
• Improvement: Maintain read ordering in reporting fusions.
• Bug fix: Update reported breakpoints so that the interval described is the last-in-read and the second is first-soft-clipped.
• Improvement: Add gene names to annotation in order they occur in reads.
• Improvement: Report all breakpoint coordinates for reads in a breakpoint cluster, add as a new tag.
• Improvement: Update read-through annotation such that we define it based on the breakpoint distance instead of the distance between genes.

Version 0.3.1

Bug fix: resolving query position sorting.

Version 0.3.0

• Bug fix: resolving strandness breakpoint errors.
• Bug fix: GTF parsing of overlapping annotations.
• Bug fix: annoting gtf records without parent gene entries.

Version 0.2.2

• Update CLI for single binary
• Add more filtering options

pbfusion v0.1.0

First release of pbfusion 🥇