Version 0.8.0 alpha 27

• feat[call]: better peak sorting
• chore[call]: revert to old SNV-finding code, but integrating candidate IDs
• fix[call]: off by one error with SNVs

Version 0.8.0 alpha 26

• feat[call]: include rsID and ref base if available for final locus SNV list
• perf[call]: optimize SNV-related code
• fix[call]: crash with no-coverage loci

Version 0.8.0 alpha 25

• chore[call]: translate contigs when using new dbSNP with RefSeq accession IDs

Version 0.8.0 alpha 24

• feat[call]: dbSNP-informed SNV calling
• feat[call]: dump callset contigs for use in MI calc
• chore[call]: don't split bootstrap iters by n of alleles when SNV-informed calling
• fix[call]: peak ordering when using SNVs
• perf[mi]: limit contig set to what is in loci dict if specified
• chore[mi]: log call loading for strkit
• fix[mi]: handle strkit peaks being None sometimes

Version 0.8.0 alpha 23

• chore[call]: require a minimum total called-in-read count when computing useful SNVs

Version 0.8.0 alpha 22

• fix[call]: handle missing snv_bases property for reads
• chore[call]: tweak a log message

Version 0.8.0 alpha 21

• chore[call]: don't look for SNVs in areas near soft-clipped regions
• chore[call]: skip SNV method when many reads are realigned to avoid bias

Version 0.8.0 alpha 20

• chore[call]: increase minimum entropy for SNV regions to 1.8

Version 0.8.0 alpha 19

• feat[call]: check entropy of area surrounding SNV to eliminate low-quality SNVs

Version 0.8.0 alpha 18

• chore[call]: tweak distance/clustering parameters for hopefully better SNV+dist clustering