NEWS
VariantFiltering 1.16
USER VISIBLE CHANGES
- Specific filtering functions have been deprecated in favor of a more general filtering mechanism.
- Shiny-app has been rewritten.
VariantFiltering 1.14
USER VISIBLE CHANGES
- The 'MafDb' class and methods have been moved to the 'GenomicScores' package.
- 'VariantFilteringResults' objects also store now a 'GenomeDescription' object that can be fetched using the method 'referenceGenome()'.
- Changes in the internal storage of 'VariantFilteringResults' objects that allow one to serialize this type of object with 'saveRDS()'.
VariantFiltering 1.12
USER VISIBLE CHANGES
- The VariantFilteringParam constructor is restricted to one (multisample) VCF file.
- mafByOverlaps() returns now a GRanges object with minor allele frequency values in the metadata columns.
VariantFiltering 1.10
USER VISIBLE CHANGES
- Updated mafById() function to speed up rs identifier look up.
- Updated scoring with weight matrices to enable using any type of weight matrix, and not only splice site matrices.
VariantFiltering 1.8
USER VISIBLE CHANGES
- Changed human defaults of VariantFilteringParam() and updated documentation on MafDb packages to reflect newer package (shorter) names of the 1000 Genomes Project.
- Sequence Ontology (SO) annotations have been updated from https://github.com/The-Sequence-Ontology/SO-Ontologies to the latest version of the data available on April, 2016.
- Analysis methods (autosomalDominant(), etc.) now accept an argument called 'svparam' that takes an object produced by the 'ScanVcfParam()' function from the VariantAnnotation package. This allows one to parametrize the way in which VCF files are read. For instance, if one wishes to analyze a specific set of genomic ranges.
- Analysis methods (autosomalDominant(), etc.) now accept an argument called 'use' that allows the user to select among three simple strategies to handle missing genotypes. See the corresponding help page for further information.
- No messages from the AnnotationDbi::select() method are given anymore about the 1:1, 1:many or many:1 results obtained when fetching annotations.
BUG FIXES
- Several bugfixes including dealing with transcript-centric annotations anchored at Ensembl gene identifiers, avoid querying for an OMIM column when working with organisms other than human, correctly identifying unaffected individuals in the autosomal recessive heterozygous analysis.
- Added methods to deal with the new ExAC MafDb packages that enable the user to query allele frequencies by position.
VariantFiltering 1.6
USER VISIBLE CHANGES
- Update on the scores() method for PhastConsDb objects that enables a 10-fold faster retrieval of mean phastCons scores over genomic intervals.
- Added two new annotated regions coded as fiveSpliceSite and threeSpliceSite and the scoring of their binding affyinity, if scoring matrices are provided.
BUG FIXES
- Fix on the scores() method for PhastConsDb objects, that was affecting multiple-nucleotide ranges from an input GRanges object with unordered sequence names.
- Fix on snpid2maf() method when decoding MafDb variants with AF values whose significant digits start with 95.
VariantFiltering 1.4
USER VISIBLE CHANGES
- Genome sequence is not fixed to a specific human genome anymore and can be parametrized via a BSgenome package
- The calculation of splice site strength on variants is now parametrized (the user can feed his/her own splice site matrices), optional and, by default, disabled because is quite computationally intensive and the user must be sure he/she wants to calculate it.
- PhastConsDb and MafDb support has been updated. Older PhastConsDb and MafDb annotation packages will likely don't work with this newer version of VariantFilering. The user must upgrade also those annotation packages.
- Analysis functions unrelatedIndividuals(), autosomalRecessiveHomozygous(), etc. now can stream over the input VCF file to reduce the memory footprint.
- Multiple single-sample VCF files are no longer supported. The user must provide always a single multi-sample VCF file.
- Internally, the package uses now the VRanges-class to store variants and they can be now also retrieved in this container.
- Indels are now annotated separtely as Insertions and Deletions. Added the annotation of Delins (deletion followed by insertion).
- Added support for annotating non-SNVs using XtraSNPlocs.* packages.
- Added support for associating BAM files to the resulting VariantFilteringResults object.
- Added some basic variant visualization capabilities via de Gviz package and the support for associated BAM files.
- Added HGVS annotations at genomic, coding and protein level.
- Added annotations to Sequence Ontology terms.
- Changed the display of VariantFilteringResults objects and added a summary() method that returns a data.frame tallying variants to feature annotations, which can be either BioC/VariantAnnotation features or Sequence Ontology features.
- Updated the example VCF files of the CEU trio to include a smaller subset of variants but called with the latest version of the GATK pipeline. Added a subset of the corresponding BAM files with aligned reads around the 1Kb region of some of the called variants.
- Integration of the IRanges/FilterRules mechanism into the filtering functionality, enabling the addition of user-specific filters.
BUG FIXES
- Some bugfixes related to matching the sequence style between variants, annotations and genome sequence.
VariantFiltering 1.2
USER VISIBLE CHANGES
- Added the calculation of the position of each variant within the cDNA (when it applies), adding also a new 'Transcript' tab in the shiny app
- Showing the number of individuals in the 'VariantFilteringParam' object, taken from the input VCF file
BUG FIXES
- Some bugfixes in the shiny app
- Replaced a .gz file in the 'extdata' folder by the corresponding .bgz file
VariantFiltering 1.0
USER VISIBLE CHANGES
- Several changes to meet requests from package reviewers.
VariantFiltering 0.99
USER VISIBLE CHANGES
- Submission of the first version to the Bioconductor project.
(start date: 26 September, 2013)