• Simplify download from GCS so fewer permissions are required
• create_coordinates script can open .gz files directly

• Bugfix for cases where start and end breakpoints get reversed (brentp/smoove#56)
• Handle INS variants in lmerge
• Allow input file paths on GCS (with gs:https:// prefixes). These files will be transparently downloaded. The user can optionally specify the temp directory to download to with --tempdir
• New svtyper with a bugfix (v0.7.1)

Major changes

• SVTyper has been updated to 0.7.0 which brings in improvements to genotyping quality, duplication sensitivity and speed improvements (courtesy of @brentp) for CRAM files.
• Support for 8-column VCFs.
• Introduction of a protocol for tiered-merging for very large cohorts via enhancements to svtools lmerge. Thanks to @abelhj.
• Multiple fixes to the classifier including manual specification of sex chromosomes which fixed a key bug when chromosomes were prefixed with a 'chr'.
• Fixes to svtools varlookup for bugs in parsing the INFO field.
• Fixed a major bug in svtools prune which occasionally resulted in incorrect results.
• Set the copynumber for the entire sample to -1 if CNVnator is unable to genotype the sample correctly. Previously svtools copynumber would die upon encountering a -1 which resulted in truncated files.
• Fixed a parsing bug in svtools vcftobedpe for BNDs on chromosomes containing a : in their name.

Minor changes

• Multiple documentation improvements.
• More helpful error handling in a number of situations.
• Better support for manta VCFs.

Major Changes

• svtyper has been updated to v0.1.2 which fixes an issue with estimating library insert sizes from CRAM files.

Major Changes

• svtyper has been updated to v0.1.1 which fixes an issue where DUP and INV variants were being genotyped incorrectly.

Major Changes

• More unified and consistent user interface, please check tool options as they may have changed.
• Moves to using v0.1.0 of SVTyper for svtools genotype. The new version is faster, uses less memory, doesn't require special tags, doesn't require a splitters file and is more accurate. Go @cc2qe!
• Major bug fix where format field definition could shuffle independent of the data fields. This only happened where the header order was different than the line's order when running svtools afreq.
• Usability improvements to svtools lsort. It can now take in a list of files (and they can be gzipped) and will skip sites where all samples are homozygous ref by default.
• Multiple improvements tosvtools prune. Fixes a bug where the variant we intended to prune was actually retained. SNAME fields are now merged during pruning so the original calls are available in the final file. Sites surviving a pruning event now receive a flag in their INFO field called RETAINED so they can be identified in downstream steps.
• Added an option to svtools lmerge to retain genotype information during merging.
• Documentation on how to install via conda
• Updates to documentation and tutorial reflecting changes and including the new svtools classify command.
• Headers are now produced even when no variants are found. (Thanks to @jasper1918)
• Numerous other fixes. Thank you to @brunogrande and @crazyhottommy for reporting issues.

Major Changes

• Changes to the BEDPE format. BEDPE now contains additional columns to aid in the reconstruction of the original VCF. These changes render the new svtools release incompatible with BEDPE generated by older versions.
• New user interface, please check tool options as they may have changed.
• Many speed improvements
• Numerous bug fixes
• Updated documentation
• svtools is now available via pip

This is a pre-release of 0.2.0 for performance testing.

This is a pre-release for testing (primarily of the release process). This software is still in alpha and not yet supported.