Fix tSVE() shiny application.
Copy parseCSQToGRanges from ensemblVEP
Substitute || by |.
tSVE vignette output format set to HTML.
parallelSlotNames() was renamed vertical_slot_names() in S4Vectors 0.25.14.
Fixes to pass updated BiocCheck requirements.
min(ranges(GRanges)) does not seem to work anymore, replaced
by more explicit expression.
Commented out some unit tests until subsetting of FilterRules
by row and column throws an error again.
Updated vignette output format for compatibility with BiocStyle (>= 2.5.19); one vignette turned to PDF to respect the package size limit in the new context.
Fixed VCF import from multiple files in Shiny application.
Update GRangesFilter following update of the condition
argument to type.
Update NAMESPACE imports following move of GRangesFilter
and GenenameFilter from AnnotationFilter
to ensembldb.
Disambiguated the variable metric that was used for two different
things in the same method and produced incorrect DataTrack name in
plotInfo.
Vignette uses alternate allele frequency to demonstrate pairsInfo
and plotInfo methods.
New argument zero.rm in method plotInfo to hide data points
with value of zero on the Y axis. Intended to reduce overplotting of
variants absent from phenotype levels, and instead emphasise variants of
low frequency.
New method pairsInfo to visualise a matrix of pairwise plots
that displays a metric calculated in levels of a given phenotype,
and stored in columns of the info slot of a VCF object.
ggpairs method imported from the GGally package.
Updated pattern used to detect INFO columns for a given metric, to be more specific.
Internal method .findInfoMetricColumns moved to utils.R,
as it is now used in two different user-visible methods.
Updated Introduction vignette to better present usage of the
addFrequencies method, better present VCF filter rules,
and introduce the new method pairsInfo.
Plotting type(s) cab be selected for the DataTrack of
plotInfo.
Update to README.
AppVeyor caches R library.
plotInfo method to visualise a metric calculated in levels of a
given phenotype, and stored in columns of the info slot.
Methods imported from Gviz package.
More methods imported from ensembldb package.
Added new VCF file and associated preprocessing script in extdata/
for gene ADH1B.
Introduction vignette updated to change activated filter rules.
Introduction vignette updated to introduce the plotInfo
method.
Moved content of the table of motivations to implement
VCF filter rules to a CSV file in misc/.
BED file and VCF files for gene ADH1B.
Updated Shell script to preprocess VCF files with the VEP script.
Ignore .svn/.
Updated reference to renamed object in Shiny app.
When no phenotypes are supplied, set phenoData slot to a
DataFrame with rownames set to colnames(vcf) and
0 columns, instead of the default behaviour of the
VariantAnnotation package which is to create a column named
Sample filled with seq_along.
NEWS file closing brackets.
autodetectGTimport setting available in tSVE method.
New checkbox in Shiny app to update selected genotypes after importing variants and autodetection of genotypes present in the data.
Do not ignore *.Rproj files.
Removed commented lines in AppVeyor YAML file.
Removed files in misc/.
Display list of error messages in a new session panel of the Shiny app.
Bumped version number to try and update the Bioconductor GitHub mirror
to display the latest code instead of version 0.1.7.
Added TVTB.Rproj to tracked files.
Deleted deprecated and misc files in inst/.
Deleted commented lines from AppVeyor settings.
The autodetectGenotypes method creates or updates
the genotypes defined in the TVTBparam that is stored
in the metadata slot of a VCF object.
The argument autodetectGT of the readVcf method may be
used to call the new autodetectGenotypes method immediately after
a VCF object is initialised from the parsed VCF file.
vepInPhenoLevel returns a GRanges instead of a
data.frame; the key advantage is that ranges may have
non-unique names.
Genotypes objects can now be initialised without specifying
ref, het, and alt genotype vectors (with a
warning).
A default Genotypes object is created with
ref, het, and alt slots set to NA_character_.
The new autodetectGenotypes method may be
used to populate those slots after variants are imported
(see New features section).
TVTBparam objects can now be initialised without supplying a
Genotypes object (with a warning).
A default Genotypes object is created (see above).
Constructors for classes Genotypes and TVTBparam are now
high-level methods, not S4 methods methods anymore.
Default settings of the Shiny app are stored as
an environment that can be overriden by arguments of the tSVE
method.
Shiny app stores more objects in reactiveValues.
Shiny app stores more error messages in reactiveValues
to better deal with optional inputs and better help users to resovle
sources of errors.
The show method throws warning messages for
TVTBparam and Genotypes objects that have not fully defined
all genotypes.
Better layout of badges in README.
Non-reactive settings of the Shiny app stored in hidden objects.
Helper methodS getEdb, tryParsePheno, tryParseBed,
tryParseVcfHeader, tryParseMultipleVcf, and
tryParseSingleVcf removed and integrated into the server side
of the Shiny app.
Massive cleaning of messages in the global.R file of the
Shiny app.
GRanges, Genotypes, and Phenotypes panels removed
from Session panel of the Shiny app.
Table reporting status of BiocParallel configurations of the
Shiny app on various system stored as an RDS file.
Shiny app displays a warning at the top of the screen if the genotypes are not fully defined.
Tab width of Shiny files set to 2.
Branches tracked by Travis CI.
Added a couple of files in inst/badexamples folder.
Added YAML file for AppVeyor.
Added pander in Suggests section of DESCRIPTION,
to render vignette tables.
README indicates status on BioC-release, BioC-devel, and Travis CI.
Updates to README: weblinks, installation, unit tests.
Branches tracked by Travis CI.
Coverage: exclude AllClasses.R, tSVE.R.
Four-space indents in DESCRIPTION.
Fixed outdated information in Introduction vignette.
Set vignette output format to BiocStyle::html_document2.
Fixed a closign bracket in NEWS file.
Replaced BiocStyle::pdf_document2() by
BiocStyle::pdf_document(); the former fails at the
pandoc
step of vignette production, the latter does not.
Fixed Collate: field of the DESCRIPTION file.
Reverted changes applied in version 0.99.3. BiocParallel does not
seem to be causing the build error on Windows Server.
Disabled BiocParallel code to see if it resolves build errors on the
Bioconductor Windows Server. Note that the documentation was not yet
updated to reflect this experimental change; this is intended to
facilitate code reversion—or alternatively document the change—in
the next commit.
Use suffix accessor in add*Frequencies methods.
New dedicated Genotypes class to store homozygote reference,
heterozygote, and homozygote alternate genotype codes,
along with the suffixes that define the INFO keys
used to store their respective data in the VCF object.
Removed families of methods tabulate* and density* from the
NAMESPACE. The features may be revisited in the future.
The associated code and documentation was saved in the inst/sandbox
subfolder for future reference.
New slot svp in TVTBparam class to store ScanVcfParam
objects. New associated accessor methods.
Moreover, TVTBparam may be coerced to ScanVcfParam.
New signatures for method readVcf that supports
param=TVTBparam, and optional phenotypes.
The method stores TVTBparam in the metadata slot of
the VCF object, and phenotypes using the colData accessor.
TVTBparam are no longer an argument of downstream methods; instead,
they must be stored in metadata(vcf)[["TVTBparam"]]
hRef and hAlt accessors renamed to refand alt,
respectively.
suffix accessor to returned named character vector
for classes Genotypes and TVTBparam.
Removed functions relevant only to the Shiny application from the
NAMESPACE (getEdb, EnsDbFilter, chr2file).
VcfFilterRules can also store instances of the parent
FilterRules class.
Defined default return value for accessors vep and type,
to avoid unnecessary switch statements.
Simplified code of inherited methods following updates to the relevant
packages (e.g. S4Vectors).
Better respect of coding standards: removed superfluous usage of
explicit argument naming, removed superfluous initialize methods.
Man pages, vignettes, unit tests and Shiny application updated to reflect changes to the package.
First release submitted to the Bioconductor review process
See DESCRIPTION file for details.