Version 0.11.0

When clustering VCF records, default is now to not put adjacent records in the same cluster (see issue #39). Rationale: gramtools graph can now handle adjacent variants, so this reduces total number of alleles by no longer generating combinations from adjacent variants.

Version 0.10.3

Bug fix handling POS<1. When the clusterer runs, it now skips these records instead of throwing an error.

Version 0.10.2

• Always print GT first (if present) in VCF line
• auto deploy releases to pypi

Version 0.10.1

• bug fix combination explosion when clustering
• tidied up code (ran black, remove pyflakes warnings)
• bug fix: clustering was not allowing POS<=1 in VCF files. Changed it so constructor does not allow POS<1 in the first place.

Version 0.10.0

Changes VcfRecord.FILTER to be a set of strings, instead of a single string, thus allowing better support for the FILTER column

v0.9.3

Add support for multi-word fasta references. Vcf record matching to the fasta reference no longer fails for multi-word fasta headers.

Version 0.9.2

Fixed bug when clustering without allele combinations. There could be the same allele sequence in the list of ALTs more than once. Duplicates are now only output once.

Version 0.9.1

Removes VCF records where the REF column does not match the reference genome.

Version 0.9.0

• add GT_CONF to genotype-aware merging
• bug fix identifying homozygous calls

Version 0.8.0

New option to merge VCF files - removes duplicates across VCF files, but does not cluster.