NEWS.md

In roryk/bcbioSinglecell: Single-Cell RNA-Seq Utilities

bcbioSingleCell 0.1.16 (2018-06-20)

New functions

• cellCountsPerCluster() and clusterCellCountsPerSample().

R Markdown templates

• Now recommending 1000 UMIs by default, and a maximum of 10% mitochondrial transcripts.
• Seurat clustering now calculates multiple resolutions, suggesting 0.4, 0.8, and 1.2 by default. Dimensional reduction plots have been updated to support looping of these multiple resolutions.

Minor changes

• `as(object, "SingleCellExperiment") coercion now slots stashed metadata into SingleCellExperiment, if defined.
• Simplified the internal code for sampleData(). seurat objects now share the same code as SingleCellExperiment, and return NULL if the sample data is not defined. The metrics() function continues to slot empty metadata in sampleID, sampleName, and interestingGroups() if not defined.
• SingleCellExperiment is used as default method over seurat where applicable: fetchData family, plotDimensionalReduction family, plotMarker family, plotFeature family, plotGene family, and plotCellTypesPerCluster(). The internal code hasn't changed, it just is defined primarily for SingleCellExperiment.
• dimRed argument has been renamed to reduction, where applicable.
• topBarcodes() can now return either a data.frame or list, containing the top barcodes grouped by sample.

bcbioSingleCell 0.1.15 (2018-06-13)

Minor changes

• Updated internal code to use text as primary argument in markdownHeader() calls.
• Updated example datasets and unit tests to match.
• Working on using pbmc4k as example dataset for vignette.

bcbioSingleCell 0.1.14 (2018-06-05)

Major changes

• Multiplexed Cell Ranger barcodes are now reformatted to be more readable and compatible with the mapCellsToSamples() utility function. This change helps simplify the internal code for cell2sample(). For example in the pbmc4k dataset, the barcode IDs are sanitized from TTTGGTTTCGCTAGCG-1 to "pbmc4k_1_TTTGGTTTCGCTAGCG". The "1" here denotes 1 sample in the matrix, which is how Cell Ranger denotes multiplexed samples in a single counts matrix. Note that the "-" character is illegal in names, so we consistently sanitize barcodes to contain "_" instead. See help("make.names") for more information on syntactically valid names in R.
• readCellRanger() no longer requires reference data defined by refdataDir, although this is still recommended.

Minor changes

• Resaved example datasets.
• Switched the example cellranger_small and seurat_small datasets to the publicly available pbmc4k dataset from 10X Genomics. Here we've subset the top 500 cells and genes by abundance. We'll use either the pbmc4k or pbmc8k dataset for the vignette in a future update.
• bcbioSingleCell() and readCellRanger() functions now consistently default to not requiring sampleMetadataFile, which is now NULL by default. For bcbioSingleCell(), if a custom sample metadata file is not provided, the function reads from the bcbio YAML metadata. For readCellRanger(), the function uses minimalSampleData() internally to return minimal metadata, containing sampleName and description columns.
• Broke out internal .sampleDirs() function to bcbioSingleCell() and readCellRanger() functions.
• Improved plotMarker documentation examples to use mitochondrial genes.
• Using seurat_small in place of Seurat::pbmc_small in working examples.

Internal code changes

• Now consistently using dplyr for piped data.frame as much as possible, where applicable. Code is being updated to use tidyeval.

bcbioSingleCell 0.1.13 (2018-05-29)

Minor changes

• t-SNE and UMAP plot improvements.

bcbioSingleCell 0.1.12 (2018-05-25)

Major changes

• No longer using automatic camel case sanitization for metrics() or fetchData return column names.
• Improved R Markdown clustering and marker templates to optionally support UMAP and dark mode in the YAML parameters.

Minor changes

• Using original Seurat mapping names for data: tSNE_1, tSNE_2, PC1, PC2, UMAP1, UMAP2.
• Ensure transcript-level counts always have stripTranscriptVersions() command applied, to remove the Ensembl transcript versions if present.
• No longer using labels (e.g. A, B, C, D) on ggplot grid return.
• Note that "phase" has been renamed to "Phase" in the R Markdown clustering for cell-cycle regression PCA.

bcbioSingleCell 0.1.11 (2018-05-23)

New functions

• UMAP is now supported. This functionality is provided in: plotUMAP(), plotMarkerUMAP(), and plotFeatureUMAP(). Corresponding fetch functions, fetchUMAPData() and fetchUMAPExpressionData(), have also been added.
• plotGene(): Added seurat method support. If advanced customization of the plot is needed, use plotDot() or plotViolin() instead, or refer to the Seurat documentation for alternates.

Major changes

• Dimensional reduction and marker plots no longer use dark mode by default. The default color palette support for marker plots has been improved to consistently use viridis.
• diffExp(): improved internal code to work directly on SingleCellExperiment, removing the need to pass design and group parameters internally. Also added unit testing against zinbwave, zingeR, and [edgeR][] support. [DESeq2][] is supported but runs slowly.
• Reworked plotFeature() and plotMarker() family of functions. Improved the color palette support when dark = FALSE, now using a flipped viridis plasma color palette.
• aggregateReplicates() function has been reworked to return a SingleCellExperiment object instead of bcbioSingleCell. The v0.2.4 update of bcbioRNASeq behaves similarly with this generic.

Minor changes

• Reworked the internal handling of some seurat SingleCellExperiment method support, using as(x, "SingleCellExperiment") internally, which uses the new Seurat::Convert() function.
• Made some previously deprecated functions now defunct: plotClusters(), plotTSNEExpressionData(), loadSingleCellRun(), darkTheme(), pcCutoff(), quantileHeatmap(), plotKnownMarkers(), readMarkers(), readMarkersFile().
• Made plotFeatures(), plotMarker(), and plotMarkers() functions defunct.
• plotPCElbow() now returns a plot grid.
• sanitizeMarkers(): improved internal code for supported bcbio stashed metadata, including rowRanges.

bcbioSingleCell 0.1.10 (2018-05-19)

Minor changes

• plotCellTypesPerCluster() is using dark = TRUE by default again.
• Fixed cell2sample() handling for multiplexed Cell Ranger data loaded up with readCellRanger(). Need to use stashed cell2sample factor saved in metadata(), rather than attempting to calculate on the fly with mapCellsToSamples().
• Updated Travis CI build checks to include bioc-release on macOS.

bcbioSingleCell 0.1.9 (2018-05-18)

Major changes

• No longer attempting to sanitize the rownames for seurat objects in coercion method. This helps maintain the gene symbol appearance in plotting functions for genes with hyphens in the names.

Minor changes

• Using BiocParallel::SerialParam() internally for zinbwave in diffExp().
• Simplified cell2sample() internal code to always use mapCellsToSamples() instead of attempting to use a stashed vector inside metadata() for SingleCellExperiment method.
• Removed internal .applyFilterCutoffs(), which is no longer necessary since this functionality is supported in the S4 subset method.
• Simplified assert checks inside fetchGene() functions.
• plotCellTypesPerCluster(): revert back to dark = TRUE by default.
• Consolidated plotMarker and plotFeature functions in the documentation.
• sanitizeMarkers(): Improved gene identifier matching.
• topMarkers() now defaults to coding = FALSE by default, since not all datasets will contain biotype information.

bcbioSingleCell 0.1.8 (2018-05-16)

Minor changes

• Initial updateObject() method support for bcbioSingleCell class.
• Relaxed validObject() validity check to not require sample-level metadata in colData() yet.

bcbioSingleCell 0.1.7 (2018-05-15)

New functions

• Added support for Uniform Manifold Approximation and Projection (UMAP) with the plotUMAP() and fetchUMAPData() functions. These work similarly to the other plotDimensionalReduction() and fetchData() functions.

Major changes

• Now adding sample-level metadata into colData() slot, for better downstream compatibility with other packages that work with SingleCellExperiment container class. Unique per-sample rows are still saved internally in the sampleData() slot.
• Now recommending ECDF as the default geom for quality control plots, where applicable.
• filterCells() now supports minUMIs = c("knee", "inflection") for automatic filtering based on the cellular barcode ranks. Internally this is handled by DropletUtils::barcodeRanks().

Minor changes

• Attempting to re-enable libgsl-dev installation for zinbwave on Travis CI.
• Suggesting BiocParallel for zinbwave call in diffExp().
• Now importing Seurat functions into NAMESPACE.
• Consolidated fetchData() functions in the documentation.
• Consolidated plotDimensionalReduction() functions in the documentation.
• Updated aggregateReplicates() internal code. This function again only supports aggregation of bcbioSingleCell objects that have been filtered using the filterCells() function.
• Now using Seurat::Convert() internally to coerce seurat class object to SingleCellExperiment, using as(seurat, "SingleCellExperiment"). This utility function was added to Seurat v2.3.1.

Internal changes

• Tweaked metrics() SingleCellExperiment method code to always merge colData() and sampleData().
• Updated readCellRanger() internal code to match bcbioSingleCell() constructor, specifically handling sample-level metadata in colData().

bcbioSingleCell 0.1.6 (2018-05-09)

Minor changes

• Updated default QC R Markdown template.
• Added trendline option to QC scatterplot functions.
• Simplified internal handling of interestingGroups in plotQC().
• Using readYAMLSampleData() internally instead of defunct sampleYAMLMetadata().
• Added sampleNames() method support for seurat.
• Now importing rmarkdown, sessioninfo, tidyverse for R Markdown reports, rather than suggesting. Similar update applied to bcbioRNASeq.
• Suggesting scater and scran.

bcbioSingleCell 0.1.5 (2018-05-04)

Major changes

• Overhauled inflection and knee point labeling support in plotUMIsPerCell(). Now uses the point argument and always labels per sample. Currently requires the geom = "ecdf" argument for labeling.
• Updated default quality control template.
• Added plotBarcodeRanks().

Minor changes

• Added barcode rank support for seurat class objects.
• QC plots now have titles by default, matching the conventions used in bcbioRNASeq.
• Fixed y-axis scale for histogram geom in QC plots.
• Prefiltering of very low quality barcodes with no UMIs or genes is now always applied. This helps avoid unwanted downstream errors with zero count barcodes.
• Added boxplot geom support for `plotReadsPerCell().
• plotQC() geom argument is now more consistent across the paneled plots.
• Fixed facet wrapping for aggregate samples in the QC plots.
• Added interestingGroups support to plotZerosVsDepth(), matching the other QC functions.

bcbioSingleCell 0.1.4 (2018-04-30)

• Updated sampleData() S4 methods to match update in bcbioBase. Now supports clean argument, which returns non-blacklisted factor columns only. See bcbioBase::metadataBlacklist for the blacklist.
• Improved axis scale appearance on dimensionality reduction plots using scales::pretty_breaks() internally.
• Added grid argument to plots, where applicable.
• Renamed example dataset from bcb_small to indrops_small.
• Removed unnecessary method support for interestingGroups() and metadata(). These extend from SummarizedExperiment correctly now.
• Fixed x-axis label centering for plotCellCounts() and plotReadsPerCell().
• Simplified seurat method support for SingleCellExperiment-like methods, where applicable. This includes rowData, gene2symbol(), and interestingGroups().
• Improved dark mode color support for plotDot(), plotFeatureTSNE(), plotMarker().
• Updated sample metadata example in README.

bcbioSingleCell 0.1.3 (2018-04-25)

Minor changes

• Improved summary statistics output during filterCells() call.
• Miscellaneous documentation improvements, most notably to bcbioSingleCell() constructor function.
• plotViolin() now uses a color border by default.
• Improved cell2sample mapping internally for readCellRanger().
• Improved Bioconductor 3.7 installation instructions.

bcbioSingleCell 0.1.2 (2018-04-24)

Major changes

• Now using bcbioSingleCell() instead of loadSingleCell() as the main constructor function to create a bcbioSingleCell object. loadSingleCell() is deprecated and still works, but will warn the user.
• Renamed loadCellRanger() to readCellRanger() for better name consistency.
• Quality control function color palettes now default to [ggplot2][] colors instead of using viridis palettes. This is defined using scale_color_hue() instead of scale_color_viridis() for example. The viridis color palette is still used by default for marker expression plots.
• Use "aggregate" instead of "sampleNameAggregate" to define aggregate/grouped samples in metadata.

Minor changes

• Reexporting relevant [ggplot2][] and viridis color palettes.
• Renamed references to "inDrops" from "inDrop", where applicable.
• Consistently using sym() in place of .data internally for tidy code.
• Updated seurat blacklist for sampleData() generic.
• plotCellTypesPerCluster() and plotMarkerTSNE() now use an automatic color palette by default, which enables for dynamic color palette support when dark = TRUE. Internally this is handled with the theme_midnight() and theme_paperwhite() [ggplot2][] themes.
• Updated installation instructions to support Bioconductor 3.7.

Internal changes

• metrics() method support now defaults to matrix and works similarly for dgCMatrix sparse matrices. This is used in place of calculateMetrics() to generate the per cell quality control metrics.
• Reworked and improved aggregateReplicates() internal code.
• Fixed facet wrapping when aggregate is defined in metadata for quality control plots.
• Improved internal code for sanitizeMarkers() to use map the gene annotations from rowRanges() better.

bcbioSingleCell 0.1.1 (2018-04-16)

Major changes

• Added support for calculating barcodeRanks() and barcodeRanksPerSample().
• Now exporting plotMarker() in addition to plotMarkers().
• Primary counts matrix slotted into assay() is named counts instead of raw, for better consistency with SingleCellExperiment class. The counts() generic requires that the primary assay slot is named counts to work correctly. Nothing else here has changed, just the name.
• loadCellRanger() now returns a SingleCellExperiment object instead of a bcbioSingleCell object.
• Added support for transgeneNames and spikeNames when loading up a dataset.
• Datasets from a poorly annotated genome can now be loaded up using organism = NULL during the loadSingleCell() call.
• Methods now dispatch on SingleCellExperiment rather than bcbioSingleCell where applicable, providing support for SingleCellExperiment objects created elsewhere.

Minor changes

• Added support for labeling barcode ranks, such as elbow or inflection point on UMI counts per cell plots.
• Added support for plotting metrics using an empirical distribution function (ECDF) plot.
• Use theme_midnight() and theme_paperwhite() internally for dimensionality reduction plots.
• TSNE and PCA plots now use an aspect ratio of 1 by default.
• Improved imports from Matrix, S4Vectors, and methods.
• Switched back to using base stop(), warning(), and message().
• inflectionPoint() has been made defunct, in favor of using barcodeRanks().
• Moved internal constructors into S4 methods, where applicable.

bcbioSingleCell 0.1.0 (2018-04-04)

Major changes

• bcbioSingleCell S4 class now extends SingleCellExperiment instead of SummarizedExperiment. This requires definition of rowRanges() inside the object instead of rowData(). Similar functionality was added to the bcbioRNASeq package. Upgrade support will be provided using updateObject()in a future release.
• Added a differential expression utility function named diffExp(), which uses zingeR/[edgeR][] internally to calculate gene expression changes across cell groups.

Minor changes

• Added plotCumulativeUMIsPerCell() utility. This may be removed in a future update in favor of adding this plot into plotUMIsPerCell() using an ECDF plot.
• Now using readCellTypeMarkers() to load marker data.frames, instead of readCellTypeMarkersFile(). This matches the conventions used in the bcbioBase package.

bcbioSingleCell 0.0.32 (2018-03-12)

• Fixes for object subsetting and loadSingleCell() organism calls

bcbioSingleCell 0.0.31 (2018-02-21)

• Improved handling of old Ensembl release version for Cell Ranger output.
• Switched R Markdown templates to consistent snake case formatting for variables and file names.
• prepareSingleCellTemplate() now uses _setup.R instead of setup.R.
• Updated unit tests to work with new assert checks.
• Export mapCellsToSamples() instead of cell2sample(). cell2sample() now simply acts as an accessor function, returning the internally stored cell2sample mappings rather than trying to calculate. mapCellsToSamples() performs the actual mapping from cellular barcodes to sample identifiers.
• Export metricsPerSample()
• Now using BiocParallel::bpmapply() to loop across the sparse matrix files per sample internally in the .sparseCountsList() function, which is shared between loadSingleCell() and loadCellRanger().
• Updated assert checks.
• Updated prepareSingleCellTemplate() to explicitly state which files to include for each R Markdown template, rather than inheriting from bcbioBase::prepareTemplate().
• selectSamples() now fails on a sample mismatch, rather than warning.
• Improved internal code for subset method, adding a drop = FALSE argument to the cellular barcode matching call.
• Made all file loads in working examples a single line.
• Simplified R Markdown directory paths.
• Made all loads in unit tests a single line.
• Renamed programs metadata slot to programVersions, to improve consistency with bcbioRNASeq package.
• Updated internal sample metadata sanitization to ensure all columns are factors.

bcbioSingleCell 0.0.30 (2018-01-26)

• Fixes for QC plot labeling of bcbioSingleCell object with different per sample filtering cutoffs applied.
• Added new fetchGeneData() function, that wraps the functionality of Seurat::FetchData() for specific genes.
• Improved internal dimensionality reduction plotting code. The main function has been renamed to .plotDR() internally.
• Simplified the code for fetchTSNEExpressionData() to return a standard data.frame with the cellular barcodes as rows, instead of the previously grouped tibble method. Now this function returns aggregate gene marker calculations in the mean, median, and sum columns. Since this method has way fewer rows than the grouped tibble, the [ggplot2][] code for plotMarkerTSNE() now runs faster.
• Explicitly declare viridis:: for color palettes, where applicable.
• colorPoints argument has been renamed to expression for plotMarkerTSNE().
• Dynamic gene symbol to ensgene conversion has been removed from the plotting functions, for greater simplicity. Now the genes argument simply matches against the rownames in the counts matrix of the object.
• Decreased the default minCumPct argument for plotPCElbow() from 0.9 to 0.8. This is more conservative and will return slightly fewer principal components for dimensionality reduction, by default.
• plotTSNE(), plotPCA(), and the other dimensionality reduction-related plotting functions now default to a smaller point size (0.5) and slight alpha transparency (0.8), to make super imposed points more obvious for large datasets with many cells.
• Added a working example for subsetPerSample().
• Internally switched from .onLoad() to .onAttach() method for automatically loading required dependency packages.
• plotPCA() now uses phase instead of Phase plotting cell cycle regression as an interestingGroup (see Seurat clustering template). Previously some of the Seurat metadata columns were not consistently sanitized to lowerCamelCase (e.g. Phase, res.0.8, orig.ident).
• Suppress package startup messages in R Markdown templates.

bcbioSingleCell 0.0.29 (2018-01-24)

• Switched to rlang methods for errors, messages, and warnings: abort(), inform(), and warn().
• Updated filterCells() function to enable per sample filtering cutoffs. This works by passing in a named numeric vector, where the names must match the internal sampleID metadata column (not sampleName).
• Improved internal sanitization of metrics available with the metrics() accessor. Now all count columns (e.g. nUMI) are consistently integers, and all character vector columns are consistently coerced to factors.
• Seurat metadata available through the bcbioSingleCell generics are now consistently sanitized in lowerCamelCase. This applies to orig.ident and the res.* metadata columns.
• Explicit integers are now consistently used in all of the function parameter arguments.

bcbioSingleCell 0.0.28 (2018-01-22)

• Manually define functions used to read barcodes and matrices. This improves the functionality of the internal .readSparseCounts() function.
• Improved sample directory matching for Cell Ranger output.
• Fixed sample metadata subsetting based on cell2sample factor levels in the internal subset code.
• Added tabbed histogram, violin, and barplots for appropriate quality control functions in the R Markdown code.

bcbioSingleCell 0.0.27 (2018-01-18)

• Migrated core dependency imports from basejump to bcbioBase.
• Improved colnames and rownames handling for internal .readSparseCounts() function.
• Reworked loadCellRanger() function. refDataDir parameter has been renamed to refdataDir.
• Added organism and genomeBuild options to loadSingleCell(), to override the metadata set in the bcbio run, if necessary.
• Improved if statement data class checks, where applicable.
• Renamed internal .sparseCountsTx2Gene() to .transcriptToGeneLevelCounts().
• Updated geomean bind method in fetchTSNEExpressionData().
• Changed minNovelty default from 0.8 to 0.75.
• Improved seurat class support for plotDot().
• Improved parameter names for plotKnownMarkersDetected(). Now uses tsneColor, violinFill, and dotColor. Also added pointsAsNumbers parameter.
• Added subtitle parameter for plotMarkerTSNE().
• Added tsneColor, violinFill, dotColor, and dark parameters for plotMarkers().
• Improved looping method for plotTopMarkers() so that it renders correctly in R Markdown calls.
• Added method support for plotViolin().
• Improved internal cell2sample handling in subset method code.

bcbioSingleCell 0.0.26 (2017-12-18)

• Renamed readMarkersFile() to readCellTypeMarkersFile().
• Improved internal handling of multiplexed CellRanger samples. These are count matrices with cellular barcodes ending in -2, for example.
• Updated aggregateReplicates() code to work with basejump generic, which uses groupings instead of cells as the grouping parameter.
• Added method support for detectOrganism().
• Improved filtering parameter output in filterCells().
• Updated gene2symbol() method support for bcbioSingleCell and seurat objects.
• Fixed working example in knownMarkersDetected().
• Reworked internal code for plotCellTypesPerCluster().
• Improved internal checks for facet wrapping and color palette parameter arguments, where applicable.
• Offloaded base plotQuantileHeatmap() functionality into basejump, for use in bcbioRNASeq package.
• Added unit test data for `loadSingleCell().
• Added additional unit tests to improve code coverage.

bcbioSingleCell 0.0.25 (2017-12-11)

• Prepared a minimal working example dataset.
• Added some initial unit tests.
• Renamed plotKnownMarkers() to plotKnownMarkersDetected().
• Renamed readMarkers() to readMarkersFile().
• Added gene2symbol() method support.
• Moved plotDot() generic to basejump package.
• Added internal .checkFormat() function, which will check for ensgene or symbol input.
• Added internal .convertGenesToSymbols() utility function, for mapping Ensembl gene identifiers to gene symbols.
• Use explicit calls to Seurat functions interally, for clarity.
• Simplified bcbioSingleCell object return in loadSingleCell() function.
• Added seurat class method support for annotable() function.
• Added bcbio<- assignment method support for seurat class objects.
• calculateMetrics() function now uses annotable = TRUE as default, instead of using missing() method.
• Added method support for cell2sample() for seurat class objects.
• cellTypesPerCluster() now uses min and max default arguments that don't remove any rows.
• Added method support for seurat class objects to counts() function. This defaults to returning the raw counts (normalized = FALSE), but can also return log-normalized (normalized = TRUE) and scaled (normalized = "scaled") counts.
• Added Ensembl gene identifier mapping support to fetchTSNEExpressionData().
• filterCells() now simply works in a destructive manner. We've removed the drop parameter. The messages displayed to the user during this function call have been improved, and now include more statistics on the step where the majority of cells are filtered.
• Initial commit of gene2symbol() method support for bcbioSingleCell and seurat class objects.
• Improved interestingGroups<- assignment method support for seurat class objects.
• Color palettes now default to viridis instead of inferno palette, where applicable.
• Added Ensembl gene identifier support for plotDot() function.
• plotFeatureTSNE() now uses a plural features parameter instead of feature, which is consistent with the syntax used in the other functions.
• Added Ensembl gene identifier support to plotMarkerTSNE(). The format argument still defaults to "symbol", for consistency with previous behavior. However, in the future we recommend that users pass in stable Ensembl gene identifiers here if possible, for better reproducibility.
• plotMarkers() now supports Ensembl gene identifiers.
• plotPCElbow() now silently returns the sequence of principal components (PCs) that we recommend to use for dimensionality reduction.
• We're now using "glm" instead of "gam" for geom_smooth() plotting, where applicable. See the plotQC() function code.
• Improved the defaults for plotQuantileHeatmap() to enable faster plotting. Now the dendrogram calculations are skipped by default, which take a long time for large datasets.
• Removed the draft plotStressGenes() function for now. Will try to add this in a future update.
• Fixed Markdown header handling for plotTopMarkers() if headerLevel = NULL.
• Simplified selectSamples() code to rely upon output of our bracket-based subsetting method. See subset.R file for more details.
• Improved metadata update in bracket-based subsetting method.
• subsetPerSample() function now defaults to saving in the working directory.
• Updated topBarcodes() function to rank by nUMI instead of nCount column, so it works with data from either bcbioSingleCell or seurat objects.
• Minor tweaks to seurat coercion from bcbioSingleCell object. Here we've improved the metadata slotting.
• Initial commit of example data script in the data-raw/ directory!

bcbioSingleCell 0.0.24 (2017-11-27)

• Raw cellular barcodes are now slotted in object@cellularBarcodes as a data.frame instead of a per sample list. This makes downstream subsetting operations on the barcodes simpler.
• Bug fixes for cell2sample mapping.
• Switched back to stable CRAN version of roxygen2 (6.0.1) for documentation.
• Renamed pcCutoff() to plotPCElbow(). The function now returns a PC sequence from 1 to the cutoff (e.g. 1:10) instead of just the final PC cutoff value. The R Markdown clustering template has been updated to reflect this change.
• Renamed quantileHeatmap() to plotQuantileHeatmap(), for consistency with other plotting functions.
• Initial support for customized bracket based subsetting, which now acts upon the raw cellular barcode counts stashed in the object@bcbio slot.
• Moved darkTheme() to basejump package and reworked as midnightTheme(), with improved colors and axis appearance.
• Added pointsAsNumbers parameter to plotTSNE() and plotPCA() functions, to match the functionality in plotMarkerTSNE().
• Overhauled loadCellRanger() to support multiplexed Cell Ranger matrix output. Cell Ranger adds a numeric suffix to the end of multiplexed barcodes (e.g. AAACCTGGTTTACTCT-1 denotes cellular barcode AAACCTGGTTTACTCT is assigned to sample 1).
• Improved cell2sample mapping in aggregateReplicates() function, which uses the sampleNameAggregate column in sample metadata to define the aggregate sample pairings. The summarize() step at line 101 is slow for datasets with many samples and should be changed in the future to speed things up.
• Improved internal cell2sample() code to handle NULL stashed mappings better.
• Updated TNSE plotting functions to use midnightTheme() instead of darkTheme().
• Added user-defined point and label sizes for plotMarkerTSNE().
• Fixed typo in plotMitoRatio() where maxGenes cutoff was plotted instead of maxMitoRatio.
• Added legend parameter argument to plotQC() function. Also improved handling of NULL return for plotReadsPerCell(), which can happen with Cell Ranger output.
• Updated facet wrapping in plotZeroesVsDepth() to match the behavior in the other plotting functions.
• Initial methods support for custom bracket-based subsetting.

bcbioSingleCell 0.0.23 (2017-11-22)

• Improved facet wrapping of aggregated samples (sampleNameAggregate present in sample metadata), but removing code support for wrapping by multiplexed FASTQ description.
• Simplified handling of bcbioSingleCell objects with filterCells() applied. This information is stored in the metadata() slot as 3 variables: (1) filterParams, numeric vector of the parameters used to define the cell filtering cutoffs; (2) filterCells, character vector of the cellular barcode IDs that passed filtering; (3) filterGenes, character vector of the Ensembl gene identifiers that have passed filtering, as determined by the minCellsPerGene parameter.
• For filterCells() return, we're now defaulting to a destructive operation, where the columns (cells) and rows (genes) of the object are adjusted to match the cells and genes that have passed filtering. Currently this can be adjusted with the drop argument for testing, but should generally be left as drop = TRUE.
• We're now slotting a cell2sample named factor in the metadata() slot, which makes downstream quality control operations faster. This is generated on the fly for previously saved objects that don't have a stashed cell2sample.
• Initial commit of plotQC() utility function, which plots multiple quality control functions for easy visualization. This defaults to output as a cowplot grid (return = "grid"), but can alternatively be set to return R Markdown code (return = "markdown").
• aggregateReplicates() operation has been improved to properly slot raw cellular barcodes in object@bcbio$cellularBarcodes. The filterCells vector is adjusted, and sampleMetadata factors should be properly releveled.
• The counts() accessor simply returns the sparse matrix contained in the assay() slot. The filterCells argument has been removed.
• Messages have been added to the filterCells() function, to help the user determine at which step the majority of cells are being filtered. We're keeping a non-destructive option using drop = FALSE for the time being, but this will likely be removed for improved simplicity in a future update.
• Updated the internal code for metrics() to use a simpler join operation on the colData, cell2sample and sampleMetadata.
• Updated facet wrap code in quality control plots to not facet multiplexed FASTQ descriptions and simply check for sampleNameAggregate.
• Improved appearance of plotReadsPerCell() labels and legends. Additionally, plotReadsPerCell() more efficiently handles the stashed values in the nCount column of colData, for faster plotting that having to rely on manipulation of the raw cellularBarcodes list stashed in object@bcbio$cellularBarcodes.
• sampleMetadata() return is now consistently sanitized for bcbioSingleCell and seurat objects.
• Minor tweaks to quality control template and setup.R files.
• Added plotFeatureTSNE() utility function. This improves on Seurat::FeaturePlot() and enables the user to overlay the cluster identifiers on top of the t-SNE plot. plotFeatures() is now deprecated in favor of this function.
• Improved internal handling of Seurat data in the .fetchDimDataSeurat() function. This now keeps the cell ID as the rowname.
• Allow the user to define the color palette (color), as well as pointSize and labelSize for plotPCA() and plotTSNE().
• Factors are now correctly releveled in cell2sample() return.
• Improved internal code for fetchTSNEExpressionData().
• Bug fix for metrics() accessor not including the cell ID as rownames.
• Clustering template fixes. Now uses plotFeatureTSNE() to assess quality control metrics on t-SNE.

bcbioSingleCell 0.0.22 (2017-11-17)

• Now internally stashing a cell2sample data.frame, which helps speed up operations on cellular barcode metrics calculations for quality control plots.
• Improved support for optional annotable, ensemblVersion, gtfFile, and sampleMetadataFile arguments in loadSingleCell() function.
• Simplified some of the messages shown during sample loading, in an attempt to make them clearer and more informative. Also improved messages shown to the user during a filterCells() function call.
• The metrics() function will now look for a stashed cell2sample data.frame, which speeds up operations for quality control plots.
• Improved handling of sample metadata columns as factors. In particular, levels should be correctly updated using droplevels in a selectSamples() call. The bcbioRNASeq package has also been updated to work in a similar fashion, where all columns in the sample metadata data.frame are now defined as factors.
• Simplified bcbioSingleCell to seurat object coercion to stash all of the bcbio metadata, and simply return the basic seurat object, rather than trying to also perform normalization and scaling. These steps have instead been added back to the Seurat R Markdown clustering template.
• Updated cell cycle and cell type markers from our master copy on Google Sheets.
• Added a troubleshooting section to the GitHub README, with a note on maximum DLLs.

bcbioSingleCell 0.0.21 (2017-11-08)

• Updated package imports to match Bioconductor 3.6.
• Initial support for plotCellTypesPerCluster().
• Initial support for plotMitoVsCoding(). I broke this code out from plotMitoRatio(). We could opt to keep this in plotMitoRatio with a geom = "scatterplot" argument.
• Initial commit of .applyFilterCutoffs() internal function, used to subset the object to contain only cells and genes that have passed quailty control filtering.
• Improved Seurat FindAllMarkers() sanitization.
• Made the quality control plots more modular. Now they support multiple geoms, including boxplot, histogram, ridgeline, and violin (default). Median labels are applied with the internal .medianLabels() function.
• Updated error message for Cell Ranger directory structure.
• Cellular barcode columns are no longer split with an underscore. Instead, they are kept as a single ACGT string. We're now generating cellID to sampleID matching with a different method. In the future, we'll stash a cell2sample data.frame inside the object, that makes this operation faster than the current mclapply() code.
• Updated annotable support in loadSingleCell() and loadCellRanger().
• Added support for handling both gene- and transcript-level counts. The updated release of the bcbio single-cell pipeline now outputs at gene level.
• Initial quality control plot support for seurat objects.
• Added support for return of only filtered cells and genes with the counts(filterCells = TRUE) function.
• Added assignment support for interestingGroups<-.
• Initial support for sample metadata generation from seurat object.
• Improved internal code for selectSamples().
• Updated topMarkers() to match Seurat v2.1 update.
• Improved bcbioSingleCell to seurat coercion method with setAs().

bcbioSingleCell 0.0.20 (2017-10-24)

• Upgraded to basejump 0.1.0 and Seurat 2.1.0 dependencies.
• Improved documentation of NAMESPACE imports per function.
• Switched to base grep functions where applicable (grepl(), gsub()).
• Use GTF in package documentation rather than GFF. Applies primarily to the loadSingleCell() import function.
• Restrict class support in S4 methods to bcbioSingleCell. Legacy bcbioSCDataSet class can be upgraded to bcbioSingleCell class using as(bcb, "bcbioSingleCell") coercion.
• Use filtered cell output for metrics() and quality control functions by default.
• Updated the quality control R Markdown to include filterCells = FALSE where applicable.
• Draft support for aggregated technical replicates in quality control functions using sampleNameAggregate column in sample metadata. This doesn't change the actual counts values. It only applies to visualization in the quality control plots currently.
• Miscellaneous R Markdown template updates. Primarily improvements to the setup chunk object loading workflow.
• Removed lintr checks from testthat. This is breaking devtools::test().

bcbioSingleCell 0.0.19 (2017-10-12)

• Renamed main object class from bcbioSCDataSet to bcbioSingleCell.
• Cell filtering with filterCells() will now slot a named logical vector into metadata(object)[["filteredCells"]], which will be used to dynamically subset the slotted internal SummarizedExperiment data. Now that we're using this approach, we can return a modified bcbioSingleCell object rather than defining a separate bcbioSCFiltered class.
• Renamed loadSingleCellRun() to loadSingleCell(), to match bcbioRNASeq package.
• Now allowing implicit integers in our function code.
• Added support for plotting technical replicates. This is handled by sampleNameAggregate in the sample metadata.
• Now using ridgeline plots in place of histograms where applicable.
• Travis CI checks take too long when loading SummarizedExperiment. Hopefully this will be fixed in the 3.6 release later this month.
• New internal dark theme (darkTheme()), based on the Seurat theme.
• Initial commit of plotDot() function, based on Seurat::DotPlot().
• Added new t-SNE plots that allow for consistent cluster labeling.
• Providing legacy support for bcbioSCDataSet and bcbioSCFiltered, which will be deprecated in a future release.
• Offloaded some internal code to basejump, for improved consistency with bcbioRNASeq package: internal-projectDir.R, internal-readSampleMetadataFile.R, internal-sampleDirs.R. We may want to provide this code as a shared bcbio core package (e.g. bcbioBase) in the future.
• Added internal utility to check for valid marker genes (.validMarkers()).
• Improved Ensembl release version support (ensemblVersion).

bcbioSingleCell 0.0.18 (2017-09-17)

• Renamed plotClusters() to plotMarkers(). Added soft deprecation.
• Added viridis color support in t-SNE plots and heatmaps.
• Converted loadSingleCellRun() and loadCellRanger() from S4 generics back to standard functions.
• Added t-SNE utility functions: fetchTSNEData(), fetchTSNEExpressionData(), and plotTSNEExpressionData(). This enable plotting of geometric mean values of desired marker genes.
• Updated NEWS to Markdown, with hyperlinks.
• Offloaded generics that would otherwise conflict with bcbioRNASeq to the basejump package.
• Improved roxygen documentation. Moved as much documentation as possible to the methods files.
• Updated cellCycleMarkers and cellTypeMarkers data. Now supports Drosophila.
• Sample IDs are now sanitized using make.names() instead of camel(). This avoids undesirable coercion of some IDs (e.g. group1_1 into group11).
• Added recommended package syntax guidelines.
• lintr checks now allow implicit integers (e.g. 1 instead of 1L).
• Added Seurat as dependency in DESCRIPTION file. The package now attaches Seurat automatically.
• Package no longer imports mononcle or suggests scater, scde, or scone. We're planning on adding these back in a future update, but build checks on Travis CI otherwise take too long.
• Added new quantileHeatmap() function.
• Improved Markdown header support across functions, where applicable.
• Improved bcbioSCFiltered to seurat coercion to slot relevant bcbio metadata.

bcbioSingleCell 0.0.17 (2017-09-03)

• Renamed package from bcbioSinglecell to bcbioSingleCell.
• Added viridis color palette support to quality control plots.
• Added cell-cycle marker genes for human and mouse.
• Slotted organism in bcbioSCDataSet metadata, in addition to genomeBuild.
• Updated bcbioSCFiltered to seurat coercion method to also run FindVariableGenes() and ScaleData() by default.
• Added cell-cycle regression into Seurat clustering [RMarkdown][] template.
• Improved pkgdown settings and website appearance.

bcbioSingleCell 0.0.16 (2017-08-25)

• Support for CRAN release of Seurat.
• Improved documentation of package NAMESPACE in bcbioSinglecell-package.R file.
• Offloaded download() functionality to basejump package, to avoid collisions with bcbioRNASeq package. Function has been renamed to externalFile().
• Removed function deprecations to simplify the NAMESPACE.
• Improved Cell Ranger sample matching. With these changes the internal .detectPipeline() function is no longer needed.
• Improved sample directory matching in internal .sampleDirs() function.
• Updated paths to Cell Ranger MatrixMarket files in internal .readSparseCounts() function.
• Updated use of packageSE() to prepareSE(), matching the corresponding basejump function change.
• Renamed internal use of filter() to tidy_filter(), to avoid future NAMESPACE collisions with ensembldb package.
• Renamed bcbioSCSubset class to bcbioSCFiltered class.
• Fixed memory issue in plotZeroesVsDepth() for datasets with high cell counts.
• Improved sample matching for selectSamples(). Will attempt to migrate this to bracket-based subsetting in a future update.
• Restricted sample selection with selectSamples() to only work on bcbioSCFiltered class for the time being. We can add bracket-based subsetting or S4 method support in selectSamples() to properly work on bcbioSCDataSet class in a future update.
• Updated [RMarkdown][] template settings and simplified the setup chunks.
• Initial support for bcbioSCFiltered class coercion to monocle CellDataSet class.
• Suggest scater and scone packages for additional quality control and visualization.
• Require monocle >= 2.5.0.
• Added initial support for viridis color palette in quality control plots.
• Suggest [rmarkdown][] and scde packages.
• Initial commit of subsetPerSample() function.
• Miscellaneous [RMarkdown][] template improvements.

bcbioSingleCell 0.0.15 (2017-08-11)

• Renamed functions in lowerCamelCase from snake_case.
• Draft support for monocle, scater, and scone.
• Package now depends on SummarizedExperiment.
• Renamed loadRun() to loadSingleCellRun() for improved compatibility with bcbioRNASeq package. This helps avoid NAMESPACE collisions between packages.
• Improved support for custom GTF files.
• Split out Cell Ranger import into a separate utility function named loadCellRanger().
• Added plotZerosVsDepth() by @roryk.
• Renamed filteringCriteria to filterParams in @metadata slot.
• Miscellaneous documentation fixes.

bcbioSingleCell 0.0.14 (2017-07-26)

• Migrated plotting functions to S4 methods.
• Improved sample metadata handling.
• Updated sample name and cellular barcode sanization.
• Initial commit of Seurat utility functions.
• Initial commit of YAML parameters for [RMarkdown][] templates.
• Draft support for Seurat v2 pre-release.
• Initial support for SureCell UMIs.

bcbioSingleCell 0.0.13 (2017-07-10)

• Integrated [bcbio][] and Cell Ranger workflows into load_run().

bcbioSingleCell 0.0.12 (2017-06-28)

• Initial support for S4 object creation using SummarizedExperiment.

bcbioSingleCell 0.0.11 (2017-06-24)

• Added testthat, lintr, and covr support for code coverage.

bcbioSingleCell 0.0.10 (2017-06-15)

• Updated [RMarkdown][] templates.

bcbioSingleCell 0.0.9 (2017-06-12)

• Compatibility update for basejump S4 NAMESPACE changes.
• Improved plots for quality control and filtering.

bcbioSingleCell 0.0.8 (2017-05-28)

• Renamed mitochondrial plot functions.
• Changed presentation of mitochondrial abundance as ratio instead of percentage.
• Simplified NAMESPACE by offloading dependencies to basejump.

bcbioSingleCell 0.0.7 (2017-05-13)

• Initial support for loading of 10x Genomics Cell Ranger output.
• Add detection of droplet method based on the metadata file.
• Draft support for import of inDrop i5 index barcode counts.

bcbioSingleCell 0.0.6 (2017-05-10)

• Modified load_run() function for improved consistency with bcbioRNASeq package.
• Initial incorporation of scater package into workflow.
• Initial commit of RMarkdown template skeletons.

bcbioSingleCell 0.0.5 (2017-05-08)

• Improvements to run object based on code in bcbioRNASeq package.
• Better handling in plots for datasets with many samples.
• Labeled quality control cutoffs on the plots.

bcbioSingleCell 0.0.4 (2017-04-26)

• Prepared package for tidyverse dependency updates.
• Updated HPC server detection.

bcbioSingleCell 0.0.3 (2017-04-20)

• NAMESPACE consolidation and function cleanup.

bcbioSingleCell 0.0.2 (2017-04-12)

• Added quality control plotting functions.
• Converted all functions to standard evaluation.

bcbioSingleCell 0.0.1 (2017-03-01)

• Initial draft release.

roryk/bcbioSinglecell documentation built on May 27, 2019, 10:44 p.m.