NEWS.md

In hbc/bcbioBase: Bcbio Base Functions

Release notes

bcbioBase 0.9.0 (2023-10-05)

Major changes:

• Bumping version due to downstream changes in Acid Genomics dependencies.
• Now enforcing strict camel case for all function names.
• Renamed bcbioBaseTestsURL to bcbioBaseTestsUrl.
• Renamed getGTFFileFromYAML to getGtfFileFromYaml.
• Renamed getMetricsFromYAML to getMetricsFromYaml.
• Renamed getSampleDataFromYAML to getSampleDataFromYaml.
• Renamed getUMITypeFromCommands to getUmiTypeFromCommands.
• Removed stringi dependency in favor of AcidBase variants.

Minor changes:

• Removed utils dependency, due to change in test URL location.

bcbioBase 0.8.2 (2023-08-15)

Minor changes:

• Now requiring R 4.3.
• Updated lintr checks and enabled testthat to run in parallel.
• Removed . from global variables.

bcbioBase 0.8.1 (2022-05-24)

Minor changes:

• Updated lintr checks and testthat unit tests.

bcbioBase 0.8.0 (2022-05-06)

Major changes:

• New release series that requires R 4.2 / Bioconductor 3.15.
• Relaxed version pinning of Acid Genomics dependencies.

bcbioBase 0.7.1 (2022-04-29)

Minor changes:

• Bumped the minimum R dependency version to 4.2.

bcbioBase 0.7.0 (2022-03-11)

Major changes:

• Removed copyToDropbox function, which is not commonly used and is difficult to properly unit test.

Minor changes:

• Updated importDataVersions and importProgramVersions to now always use base R engine, as this avoids some parsing issues that can pop up when using readr engine instead. Primarily this applies to inconsistent date formatting in data versions return, which can cause a POSIX date error to return when using readr engine.
• sampleDirs function now excludes nested pipeline directories such as bcbioRNASeq, which was added in 2021.
• Now reexporting import (from pipette), which is used in working examples.

bcbioBase 0.6.22 (2021-09-08)

Minor changes:

• Updated NAMESPACE to import some functions from methods rather than using basejump reexports: as, is, new.

bcbioBase 0.6.21 (2021-03-12)

Minor changes:

• Now importing str_match from basejump rather than stringr, reducing the number of required packages in imports.

bcbioBase 0.6.20 (2021-02-26)

Minor changes:

• Renamed "blacklist" to "denylist" internally.
• Removed reexported functions that are no longer used in other packages.

bcbioBase 0.6.19 (2021-02-21)

Minor changes:

• Updated internal YAML parser to use improved rbindToDataFrame function (see AcidPlyr and/or basejump for detail) instead of the now deprecated unlistToDataFrame approach. The new rbindToDataFrame function always returns 1:1 from nested list elements to rows, which is ideal for sample metadata and quality control metrics. Alternatively, rbindlist from data.table (with fill = TRUE) is also work a look, but it can introduce unwanted row expansion and there is no way currently to enforce 1:1 mapping. So we wrote our own function inside of basejump.

bcbioBase 0.6.18 (2021-02-19)

Minor changes:

• getSampleDataFromYAML: Fix for call to AcidPlyr::unlistToDataFrame, which now returns "name" column for nested YAML processing. See internal .sampleYAML generator for details.

bcbioBase 0.6.17 (2021-02-10)

Minor changes:

• NAMESPACE updates, following basejump v0.14 release.
• Hardened the YAML parser against nested metadata, and updated to use unlistToDataFrame function defined in AcidPlyr, rather than deprecated plyr ldply approach.

bcbioBase 0.6.16 (2020-12-03)

Minor changes:

• projectDir: Improved error message on match failure.

bcbioBase 0.6.15 (2020-10-08)

Minor changes:

• Updated package dependency version requirements.

bcbioBase 0.6.14 (2020-07-24)

Minor changes:

• Maintenance release, increasing minimum R dependency to 4.0.

bcbioBase 0.6.13 (2020-02-19)

Minor changes:

• Changed license from MIT to GPL-3.

bcbioBase 0.6.12 (2020-01-20)

Minor changes:

• Now using cli package to improve console messages.

bcbioBase 0.6.11 (2019-10-30)

Minor changes:

• Updated basejump dependencies.
• Rebuilt documentation and checked for Bioconductor 3.10 support.

bcbioBase 0.6.10 (2019-08-27)

Minor changes:

• Removed deprecated and defunct functions declared prior to v0.6.
• Deprecated read* functions in favor of import*. This naming format is more consistent with other functions used in basejump.

bcbioBase 0.6.9 (2019-08-20)

Major changes:

• Moved readSampleData and readTxGene functions to basejump. These are shared methods applicable outside of bcbio. They are still reexported here inside the package, maintaining full backward compatibility.

Minor changes:

• Reduced the number of imports and dependencies required.
• Improved handling of characters and numerics in YAML parsing.
• Lightened the package by removing dplyr and magrittr dependencies.
• copyToDropbox now looks for the optional rdrop2 package, rather than importing as a dependency. This keeps the package lighter.

bcbioBase 0.6.8 (2019-08-05)

Minor changes:

• Miscellaneous documentation improvements. Now using AcidRoxygen as the shared parameter source, following the new conventions used in basejump.
• Updated dependency versions.

bcbioBase 0.6.7 (2019-07-23)

Minor changes:

• Updated basejump dependency versions.

bcbioBase 0.6.6 (2019-07-17)

Minor changes:

• Documentation improvements to pass BiocCheck. Need to have working examples in 80% of documentation files.

bcbioBase 0.6.5 (2019-07-17)

Minor changes:

• Updated basejump dependency.
• Improved Travis CI configuration.

bcbioBase 0.6.4 (2019-05-29)

Major changes:

• Merging this release from Acid Genomics fork back to canonical HBC repo. The v0.6 release series will be used as the base for new updates to bcbioRNASeq and bcbioSingleCell packages.

Minor changes:

• Hardened user metadata input checks, particularly in readSampleData. Updated internal .isSampleData assert check function to return more informative error messages to the user on failure.

bcbioBase 0.6.3 (2019-05-05)

Major changes:

• Now pinned to R >= 3.5.

Minor changes:

• Improved Travis and AppVeyor CI configuration.

bcbioBase 0.6.2 (2019-04-22)

Minor changes:

• Improved testing to ensure compatibility with R 3.4.

bcbioBase 0.6.1 (2019-04-18)

Minor changes:

• Switched Travis CI configuration to use rnaseq Docker image.
• Updated basejump dependencies.
• Updated testthat cache to use tests.acidgenomics.com URL.

bcbioBase 0.6.0 (2019-03-28)

Updated to work with basejump v0.10 release series.

Deprecations:

• Made the deprecated bcbio ggplot2 geom functions defunct. Applies to: bcbio_geom_abline, bcbio_geom_label, bcbio_geom_label_average, bcbio_geom_label_repel.
• Additional deprecations made defunct: readLog, readTx2gene, readYAMLSampleData, readYAMLSampleMetrics.

Minor changes:

• Consolidated unit test data into Acid Genomics S3 bucket (tests.acidgenomics.com).
• Example data used for unit testing, which gets cached into tests/testthat/ are now consistently formatted using kebab case instead of snake case.

bcbioBase 0.5.14 (2019-03-22)

Minor changes:

• Updated GitHub remotes to use acidgenomics instead of steinbaugh.

bcbioBase 0.5.13 (2019-03-18)

Minor changes:

• Migrated metadataBlacklist to basejump package, for use in makeSummarizedExperiment and makeSingleCellExperiment functions. Global variable is still reexported here.
• Reduced the number of reexported functions, including pipe (%>%) and import.

bcbioBase 0.5.12 (2019-02-11)

Minor changes:

• runDate(): Updated assert check to be compatible with R 3.4 / BioC 3.6.

bcbioBase 0.5.11 (2019-01-22)

Minor changes:

• Documentation improvements.

bcbioBase 0.5.10 (2019-01-17)

Minor changes:

• sampleDirs: Added an informative message for user regarding sample name sanitization of cellular barcodes.

bcbioBase 0.5.9 (2019-01-13)

Minor changes:

• Added return value for runDate to documentation.
• Updated Travis CI and AppVeyor CI configuration.
• Added comment regarding top-level sample metadata parsing in internal YAML code.

bcbioBase 0.5.8 (2019-01-07)

Minor changes:

• Miscellaneous improvements to documentation and goalie assert checks.

bcbioBase 0.5.7 (2018-12-21)

Minor changes:

• Updated documentation to reflect new conventions used in basejump. Switch to using logical(1) instead of string, for example. This approach was inspired by the conventions used in the checkmate package.
• Switched documentation titles to use sentence case instead of title case.

bcbioBase 0.5.6 (2018-12-12)

Major changes:

• First release that has switched to using goalie internally for assert checks, in place of assertive package.
• Reworked internal YAML parsing functions and improved the assert checks.

Minor changes:

• Reorganized and removed some deprecated functions.
• Updated unit tests to reflect switch to goalie.

bcbioBase 0.5.5 (2018-11-29)

Minor changes:

• Miscellaneous documentation improvements.

bcbioBase 0.5.4 (2018-11-26)

Minor changes:

• Improve GTF file assert checks and unit testing.

bcbioBase 0.5.3 (2018-11-25)

Minor changes:

• getBarcodeCutoffFromCommands: Simplified flow in cutoff internal variable assignment.
• Reorganized deprecated functions. See deprecated.R file.
• Split out projectDir and runDate functions into separate R files, from detect.R.
• Split out YAML parsing functions into separate files from yaml.R.
• Tweaked metadataBlacklist global and added improved comments.
• Reworked .assertIsSampleData and .makeSampleData internal code.

bcbioBase 0.5.2 (2018-11-19)

Minor changes:

• Updated basejump and goalie dependencies.
• Miscellaneous documentation improvements.

bcbioBase 0.5.1 (2018-11-15)

Minor changes:

• Switched imports to simply basejump, instead of attempting to reference any basejump subpackages.
• Migrating to goalie package for internal assert checks.

bcbioBase 0.5.0 (2018-09-14)

Working towards a release candidate for Bioconductor submission.

New functions:

• projectDir: Returns the latest dated directory path inside a bcbio run. If bcbio has been run multiple times to the same upload directory, the function will return the latest project directory and warn the user.

Major changes:

• readTx2gene now returns a tx2gene class object.
• Pinning to basejump Bioconductor release candidate.
• Migrated ggplot2 geoms to basejump package.
• YAML functions now return DataFrame instead of data.frame.
• sampleDirs is now stricter about loading sample directories with names that are not valid in [R][]. This now checks for non-alphanumeric characters, including spaces, dashes, and samples that begin with a number. The bcbio pipeline will be updated to enforce these rules, to avoid unexpected downstream behavior in [R][] due to invalid names.
• Improved camel case sanitization of metrics columns in the YAML parser. Now readYAMLSampleMetrics should return percentGC instead of xGC. 5'->3' bias should return as x5x3Bias. The plyr::ldply call used to coerce from a list to data.frame will sanitize names, so we need to apply our rules before this step.

Minor changes:

• Simplified NAMESPACE imports.
• Updated deprecations.
• Linking out to recommended guidelines for development.
• Moved global lanePattern variable to basejump package.

bcbioBase 0.4.1 (2018-08-19)

Minor changes:

• prepareTemplate: Migrated function to basejump package, and simplifed to copy all files from rmarkdown/shared directory inside a package. Currently in use by bcbioRNASeq and bcbioSingleCell.

bcbioBase 0.4.0 (2018-08-08)

This is a maintenance release designed to simplify the package for long-term stability. Here we are moving all of the current S4 method support to basejump for consistency, since the methods apply to SummarizedExperiment and are not bcbio-workflow specific. We are keeping the package simple by exporting the read* family of functions here, which are designed to integrate with the bcbio output directories.

Major changes:

• Moved S4 method support to basejump: flatFiles, metrics, plotCorrelationHeatmap, plotDot, plotGene, plotHeatmap, plotQuantileHeatmap, plotViolin.
• prepareSummarizedExperiment has been moved to basejump and renamed to makeSummarizedExperiment, for consistency. The underlying code remains the same.
• Moved the SummarizedExperiment to list coercion method to basejump.
• Reduced the number of re-exported ggplot2 and viridis functions, for simplicity.

bcbioBase 0.3.2 (2018-07-31)

Minor changes:

• Offloaded separatorBar and updateMsg to basejump.
• Now using roxygen2 v6.1 for documentation.
• Miscellaneous documentation improvements.
• Simplified installation instructions, referring the user to bcbioRNASeq and bcbioSingleCell packages instead.
• lintr check fixes.

bcbioBase 0.3.1 (2018-07-24)

Minor changes:

• Added hexadecimal color support to heatmap functions.
• Improved documentation, specifying the supported types for each argument more clearly.
• Improved messages in readYAML family of functions.

bcbioBase 0.3.0 (2018-07-17)

Major changes:

• Moved some generics that provide SummarizedExperiment method suport to basejump package: gene2symbol, interestingGroups, sampleData, sampleNames, sanitizeSampleData, selectSamples, uniteInterestingGroups.
• Moved method support for SummarizedExperiment to basejump: convertGenesToSymbols, counts, gene2symbol, interestingGroups, sampleData, etc.
• Heatmap functions now only provide SummarizedExperiment method support. The documentation for these heatmap functions has been improved (and simplified), showing the supported arguments more clearly.

Minor changes:

• Now using transcriptID instead of txID in tx2gene and GRanges metadata defined in rowRanges slot of object.
• Now using aes instead of aes_string for all internal ggplot2 code, which uses tidyeval (v3.0 update).
• Moved rse_bcb and rse_bcb example data to basejump, for unit testing. These datasets are still available in bcbioBase.
• Made assertFormalAnnotationCol function defunct. No longer using to check annotationCol argument integration in heatmap functions, since we're now providing only SummarizedExperiment method support, instead of matrix method support (which is too complicated and error prone for users).
• Moved assertFormalInterestingGroups function to basejump, where the other assert check functions are defined. This helps improve package consistency.
• Updated unit tests to reflect changes in the number of exported generics.

bcbioBase 0.2.16 (2018-06-28)

Minor changes:

• Improved aggregate column handling in bcbio_geom_abline.
• Improved code coverage back to 100%.
• Reorganized R methods files to use -methods.R as a suffix, as recommended by Bioconductor, in preparation for package submission.

bcbioBase 0.2.15 (2018-06-05)

New functions:

• minimalSampleData enables easy creation of a sample metadata data.frame by simplify specifying the sample names (e.g. "description" in bcbio YAML). This function was added for easy metadata handling for 10X Cell Ranger in the bcbioSingleCell package, but is generally applicable to other bcbio datasets and may be incorporated into bcbioRNASeq in a future update.

Minor changes:

• Improved internal interestingGroups handling inside plotHeatmap family of functions.
• Removed message about transgeneNames and spikeNames in prepareSummarizedExperiment if missing transcripts are present. Now the function simply lists the genes that don't have metadata in rowRanges.
• Reorganized assert checks imports in bcbioBase-package.R file.

bcbioBase 0.2.14 (2018-05-18)

Minor changes:

• Improved formatting of package NEWS.
• Improved code coverage by adding ggplot2 unit testing.
• Added sampleData<- method support for standard data.frame.

bcbioBase 0.2.13 (2018-05-07)

Minor changes:

• Added overwrite parameter to prepareTemplate, disabled by default.
• Removed validity check for sampleData accessor.
• Improved vector return in sampleNames.

bcbioBase 0.2.12 (2018-05-08)

Minor changes:

• Migrated sanitizeSampleData from basejump to bcbioBase.
• Don't return interestingGroups column in sampleData return when clean = TRUE.
• Removed internal R Markdown shared files. These are already saved per bcbio R package.

bcbioBase 0.2.11 (2018-05-07)

New methods:

• Added SummarizedExperiment method support for sampleNames generic.

bcbioBase 0.2.10 (2018-05-03)

New functions:

• Exporting ggplot2 convenience functions for easier plotting in bcbioRNASeq and bcbioSingleCell packages: bcbio_geom_abline, bcbio_geom_label, bcbio_geom_label_average, and bcbio_geom_label_repel.

bcbioBase 0.2.9 (2018-04-30)

Minor changes:

• Improved assert check messages for interesting groups.
• Switched back to internally using make.unique instead of make.names for convertGenesToSymbols coercion on SummarizedExperiment objects. We're using make.names only when coercing seurat objects (see bcbioSingleCell code).
• Updated RangedSummarizedExperiment working example, based on bcbioRNASeq example dataset.

bcbioBase 0.2.8 (2018-04-25)

Minor changes:

• Split out assertive imports so we can pin release to bioconda.

bcbioBase 0.2.7 (2018-04-25)

Minor changes:

• Updated recommended Bioconductor installation method for 3.7 release.
• Attempt to add description to metadataBlacklist.
• readSampleData: Use description metadata column internally instead of sampleID.

bcbioBase 0.2.6 (2018-04-25)

Minor changes:

• Using make.names instead of make.unique in convertGenesToSymbols method for SummarizedExperiment class.

bcbioBase 0.2.5 (2018-04-24)

Major changes:

• sampleData now supports clean = TRUE argument, which will hide columns that contain metrics or other calculations from the user. This is useful when preparing an R Markdown report, where we only want to show relevant metadata (e.g. factor columns). This is enabled by default.
• Now using metadataBlacklist to hide specific columns in sample metadata.
• sampleMetadata is now deprecated in favor of sampleData. This function still works but will now warn the user, and should be removed from R Markdown template code.
• Added SummarizedExperiment to list coercion support, which uses the same code as flatFiles.
• readSampleData no longer requires or recommended fileName column. Only the description column is required for demultiplexed data. In the case of multiplexed samples (e.g. inDrops single-cell RNA-seq), then sampleName, index and sequence names are also required. Multiplexed cell ranger data only requires sampleName and index, since the index barcode isn't present in the counts matrix.

Minor changes:

• sampleID is no longer included in metadata priority columns. This is only to be used internally and should be hidden from the user where possible.
• prepareSampleData is no longer needed and is now defunct.

bcbioBase 0.2.4 (2018-04-22)

Minor changes:

• Reworked internal code for readYAML family of functions.
• Reorganized and improved default arguments for heatmap functions.
• readSampleData: switched to using merge instead of left_join internally.
• Reexporting viridis family of color functions, including inferno, and the British variant scale_colour_viridis.

bcbioBase 0.2.3 (2018-04-19)

• plotHeatmap, plotQuantileHeatmap: Always attempt to convert genes to symbols for heatmaps.

bcbioBase 0.2.2 (2018-04-16)

Minor changes:

• Improved unit tests for convertGenesToSymbols and gene2symbol.
• Miscellaneous documentation fixes.

bcbioBase 0.2.1 (2018-04-13)

Major changes:

• convertGenesToSymbols now has method support for SummarizedExperiment.
• Updated example datasets to use rse_bcb and rse_dds.
• Simplified internal S4 method code for heatmap functions.
• prepareSummarizedExperiment now supports transgenes with the transgeneNames argument (e.g. "EGFP"), and FASTA spike-ins with the spikeNames argument (e.g. "ERCCs"). Additionally, rowRanges and colData are no longer required and can be left NULL, although this isn't generally recommended.

Minor changes:

• Consolidated sampleYAML function code.
• Consolidated assert check imports into bcbioBase-package.R file.
• Consolidated globals into globals.R file.
• Improved internal pheatmap color support.

bcbioBase 0.2.0 (2018-03-22)

Major changes:

• Added sampleData generic. Providing method support for SummarizedExperiment class. This generic provides consistent sample metadata support in both the bcbioRNASeq and bcbioSingleCell packages. For bcbioSingleCell, colData returns information about the cells, not the samples. For bcbioRNASeq, colData and sampleData provide similar information on the samples.
• Added uniteInterestingGroups as a generic. Provides support for adding an interestingGroups column to a data frame.
• Added SummarizedExperiment method support for gene2symbol, interestingGroups generics.
• Converted prepareSummarizedExperiment from a generic to a standard function.
• prepareSummarizedExperiment now requires rowRanges and supports FASTA spike-ins with isSpike argument. This helps set up a SingleCellExperiment object in the bcbioSingleCell package.
• prepareTemplate has been converted from a generic to a standard function.
• readDataVersions and readProgramVersions now return an empty tibble if the file is missing. bcbio doesn't always output these files, so we have changed the behavior from stopping on a missing file to simply warning and returning empty.
• Added sampleDirs function, that informs the user about the names of the sample directories in the bcbio run.

Minor changes:

• Improved internal factor sanitization of sample metadata for readSampleMetadataFile and sampleYAMLMetadata.
• Exporting projectDirPattern to match bcbio project directories (e.g. 2018-01-01_rnaseq) across packages.
• Changed default of rdsToken in copyToDropbox to be NULL instead of NA.
• Reduced the number of imports from basejump package.
• Now using sessioninfo::session_info instead of devtools::session_info.
• Added AppVeyor CI support for code testing on Windows.
• Made Travis CI checking stricter and added BiocCheck.
• Simplified package dependencies in DESCRIPTION.

Deprecations:

• Deprecated sampleMetadata in favor of sampleData.
• Deprecated flatFiles in favor of using as(object, "list") coercion method instead.
• Deprecated prepareSampleMetadata in favor of prepareSampleData.
• Made annotable method on SummarizedExperiment objects defunct.

bcbioBase 0.1.4 (2018-02-20)

• Don't include sample metadata in summaryYAMLMetrics return.

bcbioBase 0.1.3 (2018-02-19)

• Now exporting assertFormalInterestingGroups in camel case.

bcbioBase 0.1.2 (2018-02-18)

• Added back checkInterestingGroups since this code is still present on bcbioRNASeq master branch.

bcbioBase 0.1.1 (2018-02-17)

• Added internal assert checks.
• Now exporting sampleMetadata<- assignment generic.
• Updated encrypted token for rdrop2 working example.
• Added assert_formal_interesting_groups assert checks. Deprecated checkInterestingGroups.

bcbioBase 0.1.0 (2018-02-13)

• Preparing version pinning for bioconda. Relaxed rlang dependency from v0.1.6 to v0.1.2 and tidyr dependency from v0.7.2 to v0.7.1.
• Added copyToDropbox function, which enables input of a list of local files and returns Dropbox paths using rdrop2.
• Updated basejump dependency to v0.2.1.
• Added assertive checks for all functions.

bcbioBase 0.0.3 (2018-01-27)

• NAMESPACE export fixes.
• Minor code cleanup to pass lintr checks.
• Switched to rlang methods for errors, messages, and warnings: abort, inform, and warn.
• Messages now consistently use backticks instead of apostrophes, as recommended by the tidyverse style guide.
• Removed tibble rownames support from prepareSampleMetadata.
• Improved code coverage to 100%.

bcbioBase 0.0.2 (2018-01-19)

• Re-export assignAndSaveData, loadData, loadRemoteData, saveData from basejump package.
• Update prepareTemplate function to use internally stored data.

bcbioBase 0.0.1 (2018-01-11)

• Initial release.

hbc/bcbioBase documentation built on Oct. 18, 2023, 1:02 p.m.