Skip to content

Latest commit

 

History

History
180 lines (139 loc) · 7.55 KB

HISTORY.md

File metadata and controls

180 lines (139 loc) · 7.55 KB
Raw

0.2.6 (17 July 2015)

  • Really fix issues with removing reference calls when using sort-vcf on structural variant calls.
  • Update clojure and associated libraries.

0.2.5 (16 July 2015)

  • Do not remove reference calls or do any chromosome renaming when sorting by position with variant-utils sort-vcf.
  • Add support for GRCh38 by avoiding issues with large numbers of contigs.
  • Add ability to only fix sample name in input VCF without doing a full prep.

0.2.4 (25 March 2015)

  • De-prioritize multi-allele calls where PL and AD values do not match a final, non-multiallelic ensemble call. Thanks to Shalabh Suman.

0.2.3 (1 February 2015)

  • Correctly handle sample names with all numbers -- parse as strings.
  • Handle BED inputs with non-system line breaks.
  • Update to BioJava 4.

0.2.2 (9 January 2015)

  • Ensemble calling: correctly handle inputs with same base filename in different directories. Thanks to Nancy Hansen.

0.2.1 (3 January 2015)

  • Do not normalize by default as part of ensemble combining. Prefer that to be done upstream by other software.

0.2.0 (17 December 2014)

  • Avoid putting fully specified indels into structural variant comparison, which can lead to masked smaller calls on heterozygous deletions. Now do exact comparisons via standard methods. Thanks to Severine Catreux.

0.1.9 (6 November 2014)

  • Handle file inputs with identical final file names (but in different directories) by uniquifying the names. Thanks to Severine Catreux.
  • Resort multisample inputs to ensemble calling with inconsistent sample orders. Thanks to Shangqian.

0.1.8 (29 October 2014)

  • Fix variant-prep problem for multiple sample input where headers would get out of sync with samples. Does not effect files previously run through GATK but would cause problems with multiple sample inputs not sorted in standard GATK sample manner.
  • Move to latest MIT licensed GATK release: 3.2
  • Library updates and fixes to work with java 1.8
  • Improve debugging output for failed lines during pre-clean step.

0.1.7 (23 July 2014)

  • Support multi-sample inputs for variant-prep, by avoiding subset to the first sample.
  • Correctly handle ensemble multi-allelic sites and posterior likelihoods, dropping PLs when they don't match the final reference call alleles.
  • Enable sorting by reference position in variant utils sort-vcf to support structural variation VCF preparation.
  • Skip realignment of very long structural variation indels during preparation. Thanks to Severine Catreux.

0.1.6 (28 April 2014)

  • Ensure Ensemble input files have consistent VCF headers for multi population variant calling. Thanks to Shalabh Suman.
  • Prioritize depth attribute when choosing representative callset from multiple choices during Ensemble calling. Thanks to Shalabh Suman.
  • Use evaluation calls instead of reference calls in concordance output. Thanks to Severine Catreux.
  • Provide comparetwo functionality to provide concordance/discordance output for a pair of variants within defined regions.
  • Avoid identifying mitochondrial-only test datasets as haploid reference genomes.

0.1.5 (15 March 2014)

  • Move to MIT licensed GATK 3.0 framework.
  • Support bgzipped inputs for variant assessment. Thanks to Severine Catreux.
  • Support lightweight loading options for gemini integration to avoid large load times with new gene tables in gemini 0.6.5.
  • Avoid running GATK VariantEval which causes intermittent java core dumps.
  • Avoid re-runs of callable regions when already prepared using non-GATK chanjo-based methods.

0.1.4 (5 March 2014)

  • Allow bgzipped/tabix inputs to validation.
  • Improve representation of ensemble variants ensuring standard expected keys are present when available in an individual genotype call. Thanks to Shalabh Suman.

0.1.3 (15 January 2014)

  • Update dependencies to the GATK 2.8.1 MIT licensed framework and associated Picard, Tribble and variant libraries. Silence phone home events from library.
  • Avoid errors on converting hg19 to GRCh37 where hg19 variants contain hg19 hap contigs with no equivalent in GRCh37. It now drops these variants instead of generating an error. Thanks to Severine Catreux.
  • Ensure BED files passed to evaluations are converted to reference material coordinates when inputs differ (hg19->GRCh37). Thanks to Severine Catreux.
  • Avoid issues with running LeftAlignVariants on indels with END tags. Thanks to Justin Johnson.
  • Move to external bcbio.run tool to help abstract out some core functionality useful in other contexts.
  • Additional flexibility for Illumina to valid VCF preparation. Allow ignoring SVs or other file types.

0.1.2 (4 December 2013)

  • Handle metrics evaluation with GATK where input calls have haploid chromosomes, in the case of mitochondrial and chrY for males.
  • Check configuration input keys for comparisons to try and provide useful error messages for unexpected keys.
  • Provide variant-utils sort-vcf, which sorts a VCF file to match a reference genome file.

0.1.1 (20 October 2013)

  • Better defaults for selecting false positives during Ensemble calling to handle common 3-caller use cases.
  • Improve output for Ensemble based calling, providing useful set INFO attribute. Thanks to Shalabh Suman for discussions.
  • Increase length requirement for exact comparisons to 100bp to account for improved resolution at longer read sizes.
  • Correctly handle overlapping structural variations of same type.
  • Improve sorting of BED files to be faster and less memory dependent. Thanks to Zhengqiu Cai for problem report.
  • Move to GATK 2.7 framework. Eliminated dependencies to reduce size of final jar.
  • Custom filtering with complex attribute comparisons, allowing usage as a commandline program.
  • Move threshold for low-coverage in discordant identification to 13 based on http://www.ncbi.nlm.nih.gov/pubmed/23773188

0.1.0 (25 August 2013)

  • Provide ensemble run target to allow easier normalization of inputs from multiple variant files.
  • Generalize filtering and combining metrics for multi-sample inputs. Allows ensemble calls on jointly called VCFs.
  • Speed improvements for normalizing input VCFs in preparation for comparison or downstream manipulation.
  • Handle merging of Illumina structural variant calls (SVs.vcf) during preparation of Illumina VCFs to GATK friendly VCFs.
  • Bug fixes for removing variants that don't match reference as part of the variant preparation. Now handle hg19 -> GRCh37 chromosome swaps and multibase deletions.
  • Bug fix for Neighboring Base Quality annotation metric for regions with no quality scores. Thanks to Zhengqiu Cai.

0.0.9 (22 June 2013)

  • Move to GATK 2.5-2 from GATK 2.3 lite. Required using new GATK framework and porting of used walkers (LeftAlignVariants) and annotators from GATK-lite code base. Removal of GATK 2.5 specific functionality which is no longer MIT licensed: recalling with UnifiedGenotyper and Variant Quality Score Recalibration.

  • Improve normalization of complex variants (MNPs and indels) with handling of additional tricky cases. Thanks to David Mittelman and Jason Wang for example cases.

  • Work towards slimming down main distribution to remove larger external dependencies. Next releases will move external functionality into own packages.

0.0.8 (6 May 2013)

  • Provide detailed breakdown of variant evaluation and grading to interface with bcbio-nextgen.

  • Fixes for normalization to handle tricky structural variation cases. Thanks to David Jenkins for examples.

  • Provide additional variant pre-cleaning steps to handle variant inputs not parseable by GATK.