CHANGES IN VERSION 1.8.0 ----------------------- NEW FEATURES o Add callGenotypes function for annotating a set of tallies with diploid genotype quality, likelihoods, etc in a way that is conformant with the gVCF spec. The methods are somewhat similar to those used by the GATK UnifiedGenotyper. CHANGES IN VERSION 1.6.0 ----------------------- NEW FEATURES o Add pileupVariants function as an alternative to tallyVariants for computing nucleotide pileups using Rsamtools, instead of gmapR. In the future, this should allow VariantTools to become independent of gmapR, but the full variant statistics will only be available via tallyVariants. USER-VISIBLE CHANGES o See the NEWS for gmapR to learn about changes to tallyVariants output. CHANGES IN VERSION 1.4.0 ----------------------- NEW FEATURES o tallyVariants will now keep ref rows if variant_strand=0; this is useful for getting information when no alts are present (e.g., for making wildtype calls). Better have a big cluster to do this over the whole genome. o add a keep_extra_stats param to TallyVariantsParam; setting this to FALSE will speed things up when the extra stats are not needed. o idVerify now supports VCF input like that output by GATK. o callableFraction() now supports GRangesList and TranscriptDb. USER-VISIBLE CHANGES o The API is now based on VRanges, a formal GRanges-derived class for representing variants; use of so-called "tally" or "variant" GRanges is deprecated. o Disable proximity filter by default; we recommend this now only for whole genome calling. o QA filtering is no longer a formal part of the calling pipeline; we recommend to apply QA filters "softly" via qaVariants() and use the results for diagnostics only. o Use BiocParallel (BPPARAM argument) for tallyVariants o VariantTallyParam deprecated; use TallyVariantsParam BUG FIXES o idVerify now correctly computes cliques instead of connected components o use the total count, rather than the ref count when calculating the alt frequency CHANGES IN VERSION 1.2.0 ----------------------- NEW FEATURES o Tally, call and export indels (using same algorithm as for SNVs). o Add post-filter that discards variants that are clumped together on the chromosome (likely mapping errors). o Add filter for masking regions like simple / low complexity repeats. o Add a filter that performs a t-test on the alt vs. ref read positions. o Add callWidtype() function for determining whether a position is variant, wildtype or uncallable, assuming the built-in variant calling filters. This is based on a power calculation that considers the coverage. o Some functions for estimating concordance between samples have been added; these were developed for sample ID verification and should be considered experimental. o matchVariants() utility for matching variants by pos and alt. USER-VISIBLE CHANGES o The VCF output is now always in expanded form (one alt per row). The AD (allele depth) geno tag contains the REF and ALT counts, while AP (allele present) indicates presence of the REF and/or ALT allele. Besides the DP tag, all other tags were removed. These changes bring VariantTools more in line with GATK. o Control alt and total counts are returned from callSampleSpecificVariants. BUG FIXES o The power cutoff in the sample-specific algorithm was not considering the minimum alt read count filter. CHANGES IN VERSION 1.0.0 ----------------------- Initial release (start date: 12 September, 2012)