2.2.3

What's Changed

• Move bystro-utils to bystrogenomics by @akotlar in #23
• fix go.mod to remove akotlar repo dep by @akotlar in #24

Full Changelog: 2.2.2...2.2.3

2.2.2

What's Changed

• switch to int8 dosages, with no null values; missing genotypes represented by -1 by @akotlar in #22

Full Changelog: 2.2.1...2.2.2

2.2.1 - Expose Schema and WriteChunks on ArrowWriter

• Expose Schema and WriteChunks on ArrowWriter
• Used for dosage-filtering in bystro repo

2.2.0 - 8bit genotype dosages

• Dosage matrices are now 8bit. This is needed for low-memory or low bandwidth systems.

2.1.2 Fix handling of no-genotype VCFs

• When supplied VCFs had no genotypes, but dosageOutput was requested, we would crash, as we attempted to read non-existent genotype columns. This release resolves that issue

2.1.1 Reduce Batch Size

• Reduces batch size to 5000 lines to reduce memory usage

Release 2.1.0

• Refactor ArrowWriter to take variadic options
• Provide --noOut argument to skip main annotationOutput

Release 2.1.0-beta2

• Refactor ArrowWriter to take variadic options
• Provide --noOut argument to skip main annotationOutput

Release 2.1.0

• Refactor ArrowWriter to take variadic options
• Provide --noOut argument to skip main annotationOutput

Release 2.0.0

This release removes the 10 allele restriction for multialelics, motivated by the dbSNP VCF datasets, which has many sites with 10+ alleles. Now multi allelic loci like:

Ref Alt
C T,A,G,CTT,CTTT,CAT,CATT,CTGT,CTTGT,CGGTC,CATCCG

will be decomposed into
C T
C A
etc

While tracking the linkage of these alleles in an optional column.

This change is considered breaking because code that relied on multiallelics consisting of more than 10 alleles being filtered will need to be updated.