search

nygenome / Conpair

Concordance and contamination estimator for tumor–normal pairs
Other
56 stars 29 forks source link

Package the tool. Support gatk4 from conda. Add --genome option. #12

Open vladsavelyev opened 6 years ago

vladsavelyev commented 6 years ago

Proposing changes that would allow distribution with bioconda and bcbio-nextgen.

  1. Packaging the tool so can be installed with ./setup.py install / ./setup.py develop, or uploaded into PyPI repository to be installed with pip. Corresponding code changes:

    • rename "modules" to "conpair" to avoid clashes with other libs on system,
    • import module files instead of sourcing with imp,
    • add setup.py,
    • move common functions to conpair/__init__.py,
    • .gitignore'd artefacts eggs/dist/build.

  2. If gatk executable is available in PATH, the tool will assume it's GATK4 and use it directly (unless --gatk or $GATK_JAR are set). It allows to use GATK4 from bioconda. (resolves https://github.com/nygenome/Conpair/issues/11).

  3. GRCh37 is assumed by default and corresponding markers files are searched, however if the users want to run against GRCh38 or the mouse genome, they can specify genome build name with -g/--genome option (as an alternative to referencing the markers file directly with --markers).