Skip to content
Commits on Feb 11, 2016
  1. Correctly specify BED file for Platypus

    Fix bug where we weren't correctly specifying BED regions since
    latest Platypus version 0.8.1 skips over files not ending with".txt" or
    ".bed". Fixes #1223
  2. Add check for BED mismatches with reference genome

    Explicitly checks for BED file contigs that aren't found in the
    reference genome, raising a helpful error. Meant to avoid confusing
    downstream problems. Fixes #1224
  3. CWL: batching of variant calls by callable regions

    With fixes to cwltool (common-workflow-language/cwltool#40) we can
    rearrange files into batches. Allows setup for variant calling by
    passing callable region files and using this to split variant calls.
Commits on Feb 10, 2016
  1. CWL: process batched inputs in parallel

    Takes batched inputs into a sub-workflow individually as a first
    step for parallelization. Does not yet handle a real run as we
    need to fix issues with file
    passing (common-workflow-language/cwltool#36) to allow BAMs and
    alignment regions as inputs to the record.
Commits on Feb 8, 2016
  1. Battenberg: add gender information to script call

    Initial work on synchronizing Battenberg with latest release. Includes
    passing gender information to script which is now required.
Commits on Feb 7, 2016
Commits on Feb 6, 2016
  1. CWL: batch into records for variant calling

    Initial steps towards support of variant calling. Add
    `batch_for_variantcall` which takes all samples and groups into
    batches (populations, tumor/normal) in preparation for variant
    analysis. We represent batches as a CWL record with each key having
    multiple values for the samples within a batch. Each record can be
    independently processed by a sub-workflow for parallelization.
Commits on Feb 4, 2016
  1. CWL: Specify resource usage for tools

    - Pull resource requests from bcbio_system.yaml file and put
      ResourceRequirement in CWL specification.
    - Pass runtime information from CWL to bcbio to make use of
      multiple cores and set memory based on system.
    - Other misc updates: docs for cloud installation with bioconda
      and bump tests to require snpEff 4.2 databases.
Commits on Feb 3, 2016
  1. Add global CNVkit scatter plot.

    - Allows comparison of CNVkit output to other tools.
    - Also fix file check issue on re-runs of priority coverage calculations.
Commits on Feb 2, 2016
  1. Merge pull request #1206 from brainstorm/master

    Better looking collectl plots. Fix Y axis for network plots to easily compare.
  2. Merge pull request #1207 from apastore/master

Commits on Feb 1, 2016
  1. CNVkit plotting: create priority BED locally

    Avoid trying to create priority BED for CNVkit scatter plotting
    in original system directory. Thanks to @ohofmann
Commits on Jan 31, 2016
  1. Docs: improve description of code development

    Describe using bcbio_python, installing within current anaconda
    directory and submitting pull requests #1203
Commits on Jan 29, 2016
  1. install: avoid certificate check with wget

    Avoids errors on older wget versions that can't follow github
    certificate redirects
Commits on Jan 28, 2016
  1. Pindel: update to parameters in Pindel-C paper

    Exact parameters not reported, but add the -C option and
    report additional events within the size range:
  2. Merge pull request #1204 from brainstorm/graph_serialize_doc

    Reverse order of xaxis and calculate deltas accordingly.
  3. CWL: re-enable quality control analysis

    Updates QC steps with new parallelization approach. Now have entire
    previously working pipeline functional with generalized inputs.
Commits on Jan 27, 2016
  1. Unify priority regions and fix CNVkit BED output

    - Only use svprioritize for priority regions instead of the undocumented
      priority_regions. Avoids two duplicated confusing options.
    - Move svprioritize regions to CNVkit plots.
    - Avoid outputs of BED and BEDgraph cnr plots for CNVkit which are
      not used and provide confusion compared to calls.
    Thanks to @ohofmann for suggestions.
  2. CWL: re-enable all sample analysis for callability

    Allow combining all samples for calculations of cross-sample
    callability. Improves approach we take for defining location of
    input binding to be more understandable.
Commits on Jan 26, 2016
  1. CWL: simplify specification of parallel regions

    Current tuple of input/output/baseline was too confusing and obscuring
    commands that were happening. Trying to simplify this by having a
    single name and then having the generation of CWL and processing
    operate on these.
  2. Skip validation if input file is empty

    RTG can fail on totally empty input files. Fixes #1183
  3. VarScan: avoid failures with empty mpileup

    mpileups would trigger a silent failure from grep, resulting in
    failed jobs without any error messages. Add ifne check in front
    to avoid this and continue. Fixes #1199
Commits on Jan 25, 2016
  1. Avoid validation errors with empty VCF inputs

    Skips checking for ROC quality metrics if fail is empty due to pysam
    errors. Fixes #1183
Commits on Jan 24, 2016
  1. CWL: re-add callability and HLA typing

    Add sample-based parallelization following split alignment, re-enables
    functionality temporarily removed while working on sub-workflows.
Commits on Jan 23, 2016
  1. CWL: fully functional parallel split alignment

    Split workflow manages taking an individual sample, indexing input
    files, splitting into chunks, aligning chunks in parallel, then
    merging back into a single sample.
Commits on Jan 22, 2016
  1. CWL: add BAM merging to alignment split

    Handles merging of BAM and HLA fastq files from a CWL sub-workflow
    splitting samples. Needs a bit of work to handle prepared files for
    output but alignment split and merge nearly functional.
Commits on Jan 21, 2016
  1. docker: system packages supporting Arvados

    Enables running using bcbio Docker images in Arvados without extra
    additions. Thanks to @tomclegg.
Commits on Jan 20, 2016
Something went wrong with that request. Please try again.