NEWS.md

BRGenomics 1.0.1 (release) & 1.1.1 (devel)

• Merge changes from 0.99.33 into release branch
• Bug workaround for getStrandedCoverage() and dependent methods for getting coverage of normalized data (apparent bug in IRanges::coverage() for weighting by normalized values)
• Bug fix in getCountsByPositions() for getting counts over an unstranded region with expand_ranges=TRUE

BRGenomics 0.99.33

• Add support for log fold-change shrinkage in getDESeqResults()
• Minor bugfix generating a warning under some conditions when non-integer binsizes are used

BRGenomics 0.99.30

• Change all ncores options to default to getOption("mc.cores", 2L)
• Rename n-dimensional binning functions aggregateByNdimBins() and densityInNdimBins()
• Redone documentation website, splitting up the vignette
• Updated package vignettes/guide pages in several places
• Other minor internal changes

BRGenomics 0.99.25

• Bug fixes with expand_ranges arguments affecting getCountsByRegions(), subsampleGRanges(), and getSpikeInNFs()
• Expanded testing, particularly for expand_ranges arguments and import_bam()
• Added options in mergeReplicates()
• Various small doc updates and minor internal changes

BRGenomics 0.99.23

• Added support for non-basepair-resolution GRanges throughout, via the expand_ranges argument
  • Substantial performance benefits for less-sparse datasets (e.g. whole read coverage)
  • Supported everywhere, including counting functions, subsampling, merging, normalization, etc.
• Rewrite of mergeGRangesData():
  • Substantial performance improvements for most datasets
  • No longer requires basepair-resolution GRanges objects
  • Added options and flexibility for merging reads as well as coverage data
• Add a mergeReplicates() function
• Rewrite of makeGRangesBRG() that significantly improves performance for sparser datasets (the datasets for which using the function makes the most sense)
• subsampleGRanges() no longer returns a basepair-resolution GRanges by default
• When field=NULL, applyNFsGRanges() no longer returns a basepair-resolution GRanges by default
• Add use_bin_numbers option to n-dimension binning functions; setting to false allows returning of bin values (the bin center) instead of the ordinal bin numbers (indexes)
• Quietly adding support for GRangesList objects (not fully tested)

BRGenomics 0.99.15

• Add pre-filtering to counting functions for performance
• Some additional clarification of readcounts vs. coverage signal in counting and import functions
• Change tidyChromosomes test
• Remove indirect links in doc pages (use only exact names of man pages)

BRGenomics 0.99.10

• Code modifications to pass Bioconductor test builds on Windows:
  • Make all examples and tests single core
  • Internally (not exported) redefine mcMap (current implementation in package parallel needs to be modified)
  • In tests and examples, test if on Windows before attempting any bigWig file import

BRGenomics 0.99.0

• Changes for Bioconductor submission:
  • Move to package versioning 0.99.x
  • Update R requirement to version 4.0
  • Add new branch R3 to allow users to install under R version >=3.5
• Various minor formatting changes to codebase

BRGenomics 0.8.1

• Fixed bug in import_bam() that produced warnings when shift argument used to shift both 5' and 3' ends of reads (i.e. when length(shift) == 2)
• Updated included external datasets to be much smaller
  • Included bam file now <200 reads
  • Included bigWig and bedGraph files derived from the bam file
• Minor update to vignette to reflect the change in the bam file
• Updated the included data() objects (PROseq and PROseq_paired)
  • Shifted PRO-seq 3' bases to remove the run-on base, and updated associated package tests
  • Added xz compression to the files
• Streamlined some method dispatch in genebodies() and getDESeqDataSet() functions

BRGenomics 0.8.0

• Add support for lists in data import functions
• Add the convenience function applyNFsGRanges()
• Significant internal changes to import_bam()
  • New test for paired-end reads (deprecated use of Rsamtools::testPairedEndBam())
  • Avoids any internal use of bpiterate()
  • Dropped dependency on GenomicFiles package

BRGenomics 0.7.10

• Add intersectByGene() and reduceByGene() functions
• Minor vignette updates

BRGenomics 0.7.8

• Substantially updated vignette
• Fully load rtracklayer (so completely exported to users)
• Add isBRG() function
• Fixed bug in spikeInNormGRanges() that failed to remove spike-in reads (aside from maintaining those reads, normalization was otherwise correct)
• Fixed minor bug in which metaSubsample() automatically added rownames with list input

BRGenomics 0.7.7

• Bug fix in aggregateByNdimensionalBins() affecting simultaneous aggregation of multiple data
• Minor updates to documentation, including an error in getDESeqResults()
• Slightly expanded vignette

BRGenomics 0.7.5

• Update bootstrapping functions
  • Add blacklisting support for metaSubsample()
  • Related to blacklisting, NA values now ignored in bootstrapping
• Add additional melt options for signal counting functions
• Further expanded support for list inputs (lists of GRanges datasets), including in getStrandedCoverage()
• Add explicit support for blacklisting in getDESeqDataSet()
• Rewrite n-dimensional binning functions, and add function for aggregating data within n-dimensional bins
  • Changed arguments in binNdimensions() to only accept dataframe inputs
  • Add densityInNdimensionalBins() function to count points in each bin
  • Add aggregateByNdimensionalBins() function to aggregate data within bins using arbitrary functions
• Added arguments for setting sample names in spike-in/normalization functions
• Various improvements and streamlining for method dispatch and flexibility

BRGenomics 0.7.0

• Added functions for counting and filtering spike-in reads
• Added functions for generating spike-in normalization factors
• Added support for lists of GRanges datasets throughout, including all signal counting functions
• Updating signal counting functions with a blacklist argument, for ignoring reads from user-supplied regions
• Added wrappers for import_bam() for several common use cases
• Update getCountsByPositions():
  • Added a melt option for returning melted dataframes
  • Now returns an error by default if multi-width regions are given (must be explicit)
• Changed argument order in getMaxPositionsBySignal()

BRGenomics 0.5.6

• Update import_bam() function
  • Added support for paired-end bam files
  • Added the shift argument
• Made metaSubsample() functions robust to unevaluated inputs
• Small performance improvement to genebodies() function
• Multicore usage is again the default for getDESeqResults()

BRGenomics 0.5.3

• Substantial performance improvement for mergeGRangesData()
• Make single-core the default for getDESeqResults()
• Fixed errant warning message in binNdimensions() for integer inputs
• Update namespace to fully import GenomicRanges and S4Vectors
• Changed R dependency, evidently required by the updated data objects
• Various documentation updates
• Added package documentation page
• Minor changes to examples

BRGenomics 0.5.0

• Increased support for multiplexed GRanges objects across all functions
• Increased performance for bootstrapping over multiple fields/multiplexed GRanges objects
• Removed requirement to set 'field' argument for gettings counts over multiplexed GRanges

BRGenomics 0.4.7

• Rewrote mergeGRangesData() to support the creation of multiplexed GRanges objects
• Made getCountsByRegions() and getCountsByPositions() to return integers if input signal is integer

BRGenomics 0.4.5

• Added and modified numerous arguments
• Increased support for normalization factors across signal counting functions
• Modified behavior of bootstrapping over multiple fields by removing explicit access to .Random.seed

BRGenomics 0.4.1

• Various documentation updates

BRGenomics 0.4.0

• Added a NEWS.md file to track changes
• Added support for bam files