peach v2.0.0

Functional:

• Significantly change formats of input and output files.
• Handle overlapping genes correctly.
• Allow overlapping variants with different ref bases (deletions and insertions at the same location, overlapping SNVs and MNVs, etc.).
• Remove requirement to include both V37 and V38 coordinates in the resource files when calling on a V37 reference genome.

Technical:

• Converted from Python to Java.

orange v3.6.0

Changes:

• Cuppa predictions with NaN likelihood are filtered in the ORANGE conversion of CUPPA results

wisp v1.2 beta

Functional:

• added Amber LOH method
• refined various aspects of somatic variant purity calculations

cuppa v2.2.0

Functional:

• Added full multi-sample support (updated Python and R components)
• Adapted CuppaDataPrep to accept generic somatic variant tsv files

Technical:

• Remove deprecated code used in CUPPA v1

bam-tools v1.3 beta

Functional:

• count overlapping bases towards standard coverage
• allow reads partially overlapping a targeted region in target-only mode

Bugs:

• handle remove reads in any contig (ie not just standard human chromosomes)
• fix tracking on flag stats for consensus reads

sage v3.4.4

Technical:

• use minimum sequence dictionary in sync fragment consensus read building. Improves performance when ref genome has a high number of alt contigs

orange v3.5.1

Bugfix:

• Workaround added for bug with mapping various ORANGE cohorts to non-existing ISOFOX cohorts.

orange v3.5.0

Changes:

• Add PurpleTranscriptImpact.reported and make PurpleVariant.reported a derived field. This uses the REPORTABLE_TRANSCRIPTS vcf field introduced in PURPLE 4.0.
• Split pharmacogenetic haplotype into separate haplotype and genotype on front page and in table.
• A new Genetic Immune Escape analysis has been added to datamodel and report. This analysis determines whether the sample uses any of the immune escape mechanisms known in cancer. See also Genetic immune landscape paper
• Technical: Removed work-around for pre v1.6 LILAC (incorrectly populating ref fragments in case of tumor-only)

amber v4.0.1

Bugs:

• ensure default min tumor depth (8) is applied in tumor/reference mode

Technical:

• handle multiple reference BAM samples in config
• handle empty sites file in PCF script

mark-dups v1.1.7

Technical:

• drop supplementaries with alignment score < 30 (to match the BWA default)