v2.2.1

Added conda.enabled = true to nextflow.config

Full Changelog: v2.2.0...v2.2.1

v2.2.0

What's Changed

• Update input table doc by @ibn-salem in #15
• Adapt SplitNCigarReads command to also split multimappers by @LKress in #17

New Contributors

• @ibn-salem made their first contribution in #15
• @LKress made their first contribution in #17

Full Changelog: v2.1.0...v2.2.0

v2.1.0

Changes

• It creates the reference genome indices if not available

Bugfix

• The SplitNCigarReads operation was failing due to a wrong parameter

v2.0.1

• Bugfix: add missing sort operation at start

v2.0.0

Changes

• Add optional step SplitNCigarReads to support RNA data
• Change Picard's MarkDuplicates by sambamba to improve performance

Release v1.9.1

Changes

• Bugfix on the sort before MarkDuplicates using default temp folder which eventually lead to out of disk space. Now it uses a local tmp folder to the nextflow working folder

Release v1.9.0

Changes

• Add an initial sort step to avoid GATK strange behaviour with unsorted BAMs
• Record software versions in the output for every sample

Release v1.8.1

Changes

• Fix issue with Java not using ad hoc temporal folder created in the Nextflow working directory. This was causing trouble in particular to MarkDuplicates

Release v1.8.0

Changes

• Upgraded GATK version to 4.2.5.0 (see release notes https://github.com/broadinstitute/gatk/releases/tag/4.2.5.0)
• MarkDuplicatesSpark which promises better performance through increased parallelization has proved unstable and thus here I downgraded to the less parallel but hopefully more robust MarkDuplicates

Release v1.7.3

Bugfix

In order to call CollectHsMetrics the workflow transforms the input BED file into the Picard's intervals format. But this was writing the intervas file in the same folder where the original BED file lied. This caused conflicts when multiple samples ran at the same time and tried to write/read from/to the same file.