v0.6.0

Patch

• Fix indentation bug in quant.py LOH functionality by @jbreynier in #88
• Conda environment and dependency patch by @abuendia in #120
• Dockerfile and CI/CD to use conda environment by @abuendia in #121
• fix key error by @alienzj in #94
• Convert to BAM directly while sorting by @andreas-wilm in #78
• Update dependencies for bioconda integration by @abuendia in #123

v0.5.0

Added

• Extended arcasHLA quant module with additional --LOH option which enables loss of heterozygosity (LOH) calling as a last step provided user inputs values for purity and ploidy (per our publications). Additional flags --purity and --ploidy provided to allow user to specify estimated values for each.

Fixed

• Fixed issues causing arcasHLA quant to crash immediately when called on single-end reads.

v0.4.0

Added

• arcasHLA customize and arcasHLA quant modules added along with necessary reference files, providing new functionality for quantifying HLA allele-specific expression (per our publications). This functionality has been merged from the orginal arcasHLA-quant repository.

Changed

• arcasHLA now requires the latest version of Git Large File Storage to download this repository due to the large size of reference FASTA files.
• Both arcasHLA customize and arcasHLA quant updated to use JSON reference files instead of Python's binary pickle format.

Fixed

• arcasHLA quant now uses appropriate defaults for fragment length and standard deviation when run on single-end reads.
• arcasHLA customize now works from outside of the arcasHLA directory.
• arcasHLA merge properly processes JSON output files.

v0.3.0

Changed

• --version flag for reference now accepts versions 3.37.0-3.46.0 (arcasHLA requires Git Large File Storage to properly download later IMGT/HLA references).

Fixed

• --commit flag for reference script no longer builds index from latest commit of the IMGT/HLA reference before building from the intended commit of the IMGT/HLA reference when run for the first time.

v0.2.5

Changed

• arcasHLA reference files are now stored as human-readable JSON files instead of Python's binary pickle format.
• arcasHLA now uses defaults for fragment length and standard deviation when run on single-end reads.

Fixed

• arcasHLA reference module has been updated and functionality restored. Consistent with v0.2.0 release notes, due to the file sizes of the data stored in the IMGT/HLA repository, arcasHLA requires the latest version of Git Large File Storage to download later IMGT/HLA references. However, due to the depreciation of older git-lfs commands, there were newly introduced logical errors which have been resolved with this release.

v0.2.0

Added

• Convert module that allows for conversion of HLA nomenclature to P-groups and G-groups and reduction of allele resolution.

Changed

• arcasHLA now requires Git Large File Storage to properly download later IMGT/HLA references.
• Merge module will now combine HLA locus read counts and relative abundances into a single file.
• A minimum read count in order to type variable has been added genotype and partial modules.
• reference --version now accepts versions 3.3.0-3.36.0

Fixed

• Fixed empty priors in genotyping.
• Properly deletes temporary folders.

v0.1.0

Added

• Random names for temp folders.
• Required versions for dependencies in README.md.

Changed

• Stderr output from system calls (e.g. samtools sort) piped to log.
• Modifications so arcasHLA can be run from any location (Pull Request #2).
• Merged shared alignment scripts for genotype and extract into align.py.
• Modified architecture of aligment.p files to support merged scripts (still compatible with v0.0 alignment.p files).

Fixed

• Genotype compatibility with .fastq.gz and .fastq files (Issue: Fastq as input #7).
• Catch corrupted or empty hla.dat files (Issue: Error in ./arcasHLA reference #4).

First release w/ minor fixes.

Changed

• Changed version number to comply with semantic versioning.
• Updated test example to include partial typing.

Fixed

• Fixed issue with partial module (Issue #1).