Skip to content


Subversion checkout URL

You can clone with
Download ZIP
Commits on Nov 15, 2015
  1. Merge branch 'nh_add_max_coverage' of

    … into nh13-nh_add_max_coverage
    Also add info about the new option to parameters.
  2. Merge pull request #216 from ekg/gvcf-chunk

    gVCF chunk
  3. Merge pull request #153 from ryan-williams/patch-1

    fix some README typos
  4. Merge pull request #211 from travc/unique-bam-file-check

    duplicate bam file check
  5. Merge pull request #209 from travc/master

    bit stricter BED processing
  6. add test for gVCF chunking

  7. Merge branch 'master' of

  8. Merge pull request #215 from dnanexus-rnd/mlin-gvcf-segfault

    gvcf bugfixes
  9. adds gvcf chunking

    Untested in this commit. I need to merge @mlin's changes.
Commits on Nov 14, 2015
  1. @mlin
  2. @mlin
  3. Setting Release-Version v1.0.0

  4. Merge pull request #213 from ekg/gvcf-support

    add gVCF support
  5. add gVCF support

    When --gvcf is supplied, freebayes will output blocks for each
    region that does not have a call. These have a particular format
    that is not exactly the same as other records. A genotype quality
    is provided, but this is just the reference quality sum adjusted
    by the alternate quality sum. These are provided per-sample.
    Average depth is provided across the region. Improvement of this
    format is essential, and this version should stand as a
    placeholder for future development!
  6. @travc
  7. @travc
  8. @travc
  9. @travc

    Check for duplicate bam files

    travc committed
Commits on Nov 13, 2015
  1. @travc
Commits on Nov 11, 2015
  1. Merge pull request #186 from mckinsel/fix_bash_sp

    Fix freebayes-parallel word splitting
  2. @mckinsel

    Prevent word splitting in freebayes-parallel args

    mckinsel committed
    You might want to call freebayes-parallel with filenames with spaces
    in them:
    freebayes-parallel 'a regions file.txt' 2 -f 'some reads.fasta' alns.bam
    But, when the argument array is assigned to the command string, everything
    gets split on spaces, and freebayes would do something like complain about
    not being able to find a file called "some". This change keeps the arguments
    in an array and prevents splitting of the regionsfile.
    Two tests are included with filenames that have spaces in them.
Commits on Oct 23, 2015
  1. resolves #206

Commits on Oct 16, 2015
Commits on Sep 30, 2015
  1. resolves #188 by simplifying alleleparser position stepping

    This disables a (buggy) optimization where the parser would jump to the next
    position with alignments or input alleles. This should have minimal effect on
    performance at a benefit of greatly decreasing the complexity of the position
    increment process.
Commits on Sep 24, 2015
  1. @nh13
Commits on Sep 18, 2015
  1. Merge pull request #196 from mckinsel/fix_variantinput

    Fix variant input when region is specified
Commits on Sep 17, 2015
  1. @mckinsel

    Fix variant inputs when region is specified

    mckinsel committed
    When a variant input file is given with -@,
    AlleleParser::usingVariantInputAlleles is set to true. This makes AlleleParser
    look for variants in AlleleParser::inputVariantAlleles. But this is only
    populated if AlleleParser::targets is empty, which is only true if a region
    is not specified. This change populates the inputVariantAlleles map even when
    targets is not empty.
  2. @mckinsel

    Add variant input tests with -r and -c

    mckinsel committed
    Presently, if you set a region with -r and a variants input file with -@, the
    variant inputs will be ignored. These tests confirm that, and when they pass
    will confirm a fix. In addition, the method responsible for this behavior also
    does special things when reading the BAM from stdin, so this change adds some
    test coverage for that as well.
  3. @mckinsel

    Add tests for variant input

    mckinsel committed
    The current test for the -@ parameter is this:
    is $(freebayes -f tiny/q.fa -@ tiny/q.vcf.gz tiny/NA12878.chr22.tiny.bam | grep -v "^#" | wc -l) 19
    but this also passes without the variant input flag:
    is $(freebayes -f tiny/q.fa tiny/NA12878.chr22.tiny.bam | grep -v "^#" | wc -l) 19
    So, it's not really constraining the behavior of -@ very much. This commit
    adds two test that use a vcf with fake variants at q:500, q:1000, and q:11000
    and checks that they appear in the output vcf or that only they appear in the
    output vcf when -l is used.
Commits on Jun 5, 2015
  1. @AlistairNWard

    Update Makefile

    AlistairNWard committed
Commits on May 21, 2015
  1. correctly handle --variant-input

Commits on May 20, 2015
  1. resolves #165

    I've verified that freebayes now only works using whole-chromosome sequences.
    The anti-pattern of trying to cache little bits of the reference has been
    resolved by removing offending methods in the AlleleParser. The system should
    also run much faster.
    This likely affects #163, #166, #168, #170, but reporters need to test before I
    close these.
Something went wrong with that request. Please try again.