Arriba v2.2.0

• improved detection of internal tandem duplications
• better sensitivity for the detection of viral integration sites
• inclusion of additional ~4500 viruses into screening, including rare strains of cancer-associated viruses (requires rebuild of STAR index)
• viral contigs were renamed to be compliant with the SAM format specification (requires rebuild of STAR index)
• support for mm39/GRCm39
• utility scripts (see also manual):
  • quantify virus expression
  • convert Arriba's custom output format to VCF
  • extract fusion-supporting alignments into separate mini-BAM
  • running Arriba on a prealigned BAM file and realigning only the fusion candidate reads saves ~80% of the CPU time compared to a complete realignment (useful when the alignments were generated by an old STAR version or by a different aligner such as HISAT2)
• polishing of fusion visualizations created by draw_fusions.R and new features:
  • all transcripts can be drawn at the same scale if desired (--fixedScale)
  • circos plots have same size across all pages
  • set PDF title and print as header on every page (--sampleName)
  • fine-grained control over region to draw for intergenic breakpoints (--showIntergenicVicinity)
  • choose a different font (--fontFamily)
  • better scaling for coverage track
• more fixes for prediction of reading frame
• better warnings and error messages
• updated STAR to version 2.7.10a, which fixes malformed chimeric alignments for paired-end reads with small insert size
• updated dependencies (HTSlib, libdeflate)