- Fix bug in 1.0.0 release with parallel calculations on whole genome samples.
The release version only parallelizes by chromosome instead of callable
regions, resulting in less parallelism. Thanks to Sven-Eric Schelhorn and
Neill Gibson.
- Generalize use of working directories to support runs on S3 mounted
filesystems. Ensures all work takes place inside transactional directories.
Thanks to Tetiana Khotiainsteva and Sven-Eric Schelhorn.
- Provide separate germline calling for somatic tumor/normal pairs. Supplements
somatic calls with standard germline calls on normal samples, including
ensemble and SV calling.
- Support creating GEMINI databases with new generic mechanism using vcfanno/vcf2db.
This allows creation of GEMINI output for any organism. Adds support for hg38
with annotations from dbSNP, Clinvar, ExAC and ESP.
- Support FreeBayes 1.1.0 for improved memory usage and 3-4x speedup.
Will default to 1.1.0 at next release. Validation work:
https://github.com/bcbio/bcbio.github.io/blob/master/_posts/2016-11-21-giab-hg38-freebayes.md
- Rework quality control for speed and output directory consistency. Avoid
re-duplicating calculations and put all output in qc directory to make re-runs
easier. Thanks to Vlad Saveliev.
- Fixes for Seq2C concurrency problems when preparing BED files. Thanks to Vlad
Saveliev.
- Update WHAM structural variant caller to support the latest release.
- Update delly structural variant caller to support the latest release.
- Improve dbSNP annotation speeds for adding rs IDs to VarDict output.
Thanks to Ben Liesfeld.
- Support for VEP 87 with additional plugins and generalization of fields.
Thanks to Matthias De Smet.
- Deprecate `clinical_reporting` parameter and introduce new
`effects_transcripts` parameter than enables more control over variant effects
prediction. Enable HGVS by default for human projects and separates from
transcript selection.
- For lumpy runs that use samblaster, use samtools sort instead of sambamba
sort. Avoids segfault issues with samblaster. Thanks to Oliver Hofmann.
- Pre-install capture region BED files and enable short hand specification in
sample configuration.
- Use vt normalize as part of GEMINI decomposition to clean up complex
multiallelic variants. Thanks to Sergey Naumenko.
- Testing suite cleanup. Move to py.test and separate integration and unit
tests. Thanks to Tetiana Khotiainsteva.
- Fix issue with cutadapt hanging on gzipped input. Thanks to Stephen Turner.
- Updated cutadapt to use single-pass trimming for paired-end files, improving
performance and hitting the disk less.
- Added support for cellular barcode error correction with single-cell RNA-seq
via the `cellular_barcode_correction` parameter. This corrects edit distances
up to the set value, defaults to 1.
- Add support for sample-based demultiplexing of single-cell RNA-seq runs.
- Move single-cell RNA-seq results to the upload directory.
- Make positional UMI default to off for single-cell RNA-seq.
- Add support for the Klein lab v3 version of the inDrop protocol.