NEWS.md

In steinbaugh/pointillism: Single-cell RNA-seq toolkit

Release notes

pointillism 0.7.0 (2023-10-04)

Major changes:

• Now requiring R 4.3 / Bioconductor 3.17.
• Added initial support for pending Seurat 5 release.
• Now enforcing strict camel case for all function names.

pointillism 0.6.2 (2022-10-24)

Minor changes:

• Hardened dependency version checks.
• Updated roxygen2 documentation.
• Updated markdown files to pass lintr checks.

pointillism 0.6.1 (2022-06-09)

Minor changes:

• Updated lintr checks and testthat unit tests.
• Migrated example Seurat and SeuratMarkersPerCluster objects back to AcidTest package.

pointillism 0.6.0 (2022-05-11)

Major changes:

• Now requiring R 4.2 / Bioconductor 3.15.
• Split out basejump dependencies into individual packages.
• runSeurat: Simplified processing of TSNE and UMAP.
• Updated unit tests and working examples.

Minor changes:

• Slotting utils::sessionInfo instead of sessioninfo::session_info, where applicable.
• Removed dependency on edgeR, and removed dplyr from suggested packages.

pointillism 0.5.0 (2021-03-03)

Major changes:

• Simplified NAMESPACE, incorporating changes in basejump v0.14 release series.
• Hardened Seurat to SingleCellExperiment interconversion methods and unit tests.
• plotStackedBarPlot: Removed internal dplyr dependency.

pointillism 0.4.14 (2020-10-12)

Minor changes:

• Renamed alpha argument to alphaThreshold, matching naming conventions used in other Acid Genomics packages. Applies to marker class objects.
• Updated dependencies to support naming changes in Acid Genomics packages.

pointillism 0.4.13 (2020-08-04)

Minor changes:

• Removed patrick dependency for unit tests.
• Updated other package dependency versions.
• Wrapped suppression calls with { where necessary.

pointillism 0.4.12 (2020-07-23)

Minor changes:

• Maintenance release, updating R dependency to 4.0.

pointillism 0.4.11 (2020-06-27)

Major changes:

• runSeurat: Changed umapMethod default from "umap-learn" to "uwot". I'm having trouble getting the latest version of umap-learn (0.4.4) to play nicely with Seurat (3.1.5) via reticulate (1.16). Note that additional internal arguments are now passed to Seurat::RunUMAP when necessary.

pointillism 0.4.10 (2020-06-11)

New functions:

• plotStackedBarPlot: Added support for quickly generating stacked bar plots of cluster frequencies.

Minor changes:

• Dependency updates.

pointillism 0.4.9 (2020-02-21)

Minor changes:

• runZinbwave: Updated internal zinbwave code to be compatible with Bioconductor 3.11 release.
• Updated test data.

pointillism 0.4.8 (2020-02-20)

Major changes:

• plotReducedDim, plotMarker: Changed title handling support. Switched from using title argument in favor of labels, which better matches the conventions now used in AcidPlots.

Minor changes:

• Seurat methods now show resolution parameter used to map idents, where applicable. Currently this applies to plotReducedDim family of functions.

pointillism 0.4.7 (2020-01-30)

Minor changes:

• Now using cli package for messages.
• Updated basejump dependencies.

pointillism 0.4.6 (2020-01-20)

Minor changes:

• Reworked NAMESPACE to support migration of bioverbs package to AcidGenerics.

pointillism 0.4.5 (2020-01-03)

Major changes:

• normalize: reworked internal code to reflect changes in scater package. Instead of calling normalizeSCE or normalize internally, normalizeCounts is the recommended method for returning normcounts and logcounts assays that can be slotted in assays of the SingleCellExperiment object.
• Deprecated plotDot in favor of plotDots (note plural spelling change). This reads better and improves consistency with conventions used in scater.

Minor changes:

• Removed user-defined option to reduce verbosity, where applicable. If these functions are too noisy, we'll tone it down in the future.

pointillism 0.4.4 (2019-10-30)

Bioconductor 3.10 compatibility fixes.

pointillism 0.4.3 (2019-10-26)

Minor changes:

• Bug fix for legacy Seurat objects created via pointillism with named rownames containing gene identifiers mapped to the gene symbols. Now the coercion method for Seurat to SingleCellExperiment checks for this and sanitizes the names automatically so the resulting SingleCellExperiment is valid.

pointillism 0.4.2 (2019-09-04)

Major changes:

• Reworked internal code to remove dplyr, magrittr, rlang, tibble, and tidyr dependencies. Consistently using base R and Bioconductor S4 methods internally instead, working primarily with DataFrame instead of tbl_df.

pointillism 0.4.1 (2019-08-12)

Minor changes:

• Bug fix for gene symbols being sanitized by as.data.frame or as_tibble coercion inconsitently across R versions. Switched to using as coercion from DataFrame to data.frame instead.
• Improved message consistency, removing use of backticks in favor of single quotes, which is now the convention across basejump.
• Lintr check fixes.

pointillism 0.4.0 (2019-08-07)

Initial support for monocle3 cell_data_set class.

Major changes:

• Reworked internal S4 function names.
• Updated internal example data.
• Moved cell-cycle and cell-type markers from Google Sheets here into the package internally at inst/extdata/markers/.
• Working on preliminary support for monocle3 cell_data_set.

pointillism 0.3.2 (2019-04-25)

Minor changes:

• S4 generic reexport documentation fixes.

pointillism 0.3.1 (2019-04-23)

Minor changes:

• Switch to importing ggplot2 code from new AcidPlots package. Recommended default ggplot2 themes are now named acid_theme_light and acid_theme_dark.
• Consolidated S4 class and generator function documentation into single Rd files, where applicable.

pointillism 0.3.0 (2019-04-16)

Major changes:

• Updated S4 methods to support Seurat v3 release. Also updated imports from Seurat to reflect v3 change: GetAssayData, Stdev, VariableFeatures.
• Now importing S3 coercion methods from Seurat: as.SingleCellExperiment. as.Seurat, which coerces SingleCellExperiment, isn't particularly helpful because it requires logcounts slot to be defined in SCE. We're calling CreateSeuratObject internally in our S4 coercion method instead, so that any SCE object can be easily coerced to a Seurat object.
• Resaved and renamed example data: seurat, seurat_all_markers, seurat_known_markers.
• Overhauled R Markdown templates.

Minor changes:

• Added preliminary plotCellTypesPerCluster method support. Need to test this out on larger datasets with many samples per cluster.
• Switched Travis CI configuration to use Docker image.
• Added some additional comments regarding reticulate umap-learn configuration required to run Seurat::RunUMAP. Consider putting some information in the README, since there is a lot of misinformation on GitHub regarding proper UMAP configuration.
• Removed setOldClass definitions for session_info and package_version.
• Renamed as-methods.R to coerce-methods.R, matching bsaejump conventions.
• cellCountsPerCluster: Improved internal interestingGroups handling.
• Reworked reexport method for S4 generics imported from bioverbs package.

pointillism 0.2.5 (2019-04-01)

Major changes:

• Migrated code to Acid Genomics.
• Made runZinbwave defunct. No longer recommending zinbwave calculations for droplet single-cell RNA-seq.
• ggplot2 themes theme_paperwhite and theme_midnight are now imported from minimalism package instead of basejump.

pointillism 0.2.4 (2019-01-24)

Minor changes:

• SeuratMarkersPerCluster to tbl_df coerce needed a decode call to properly handling gene-to-symbol mappings with run-length encoding (Rle).

pointillism 0.2.3 (2019-01-08)

Minor changes:

• Comment out lines to pass lintr checks.
• Split out imports into a separate imports.R file.
• Added nullOK flag in assert checks, where applicable. See isGGScale, for example.

pointillism 0.2.2 (2018-12-22)

Major changes:

• Migrated S4 generics to bioverbs package.
• Finish migrating assert checks to goalie package.

Minor changes:

• Moved S4 validity checks into setClass call, rather than using a separate setValidity call. Refer to AllClasses.R for details.
• Removed makeCellTypeMarkersFromGoogle utility function.

pointillism 0.2.1 (2018-12-01)

Major changes:

• Reworked internal code to use goalie package for assert checks.
• SeuratMarkersPerCluster: Simplified S4 validity checks.
• Improved show methods for S4 classes: CellCycleMarkers, CellTypeMarkers.
• Added summary method support for SeuratMarkers, SeuratMarkersPerCluster.

Minor changes:

• Miscellaneous documentation improvements and fixes.
• Reworked internal code shared between CellCycleMarkers and CellTypeMarkers S4 generator functions.
• SeuratMarkers: Overhauled internal code, moving from .seuratMarkers internal function. Refer to SeuratMarkers.R for details.
• Added AppVeyor CI support.

pointillism 0.2.0 (2018-11-15)

New functions:

• CellCycleMarkers, CellTypeMarkers, and KnownMarkers generator functions.
• cellTypesPerCluster S4 generic.
• topMarkers S4 generic.
• New functions to import sheets from Google: importCellCycleMarkersFromGoogle and importCellTypeMarkersFromGoogle. These may be removed in a future update, in favor of simply using CSV files managed in the package. We've also included a helpful makeCellTypeMarkersFromGoogle utility function.
• SeuratMarkers: New function for returning sanitized Seurat marker output.
• SeuratMarkersPerCluster: Returns an S4 class split by cluster.

Major changes:

• Improved S4 coercion methods. Added support for CellCycleMarkers to tbl_df, CellTypeMarkers to tbl_df, and SeuratMarkersPerCluster to tbl_df.
• Reworked internal code for plotFeature.
• plotKnownMarkers is now exported instead of plotKnownMarkersDetected.
• Improved formals (and consistency) for dimensional reduction plotting functions. This applies to plotReducedDim, plotTSNE, plotPCA, and plotUMAP.
• runZinbwave: Improved internal code and weight handling.

Minor changes:

• Reworked validity checks for S4 classes.
• Reworked internal assert checks. Refer to .assertIsBPPARAM, for example.
• Broke out S4 methods into internal functions, where applicable. Refer to cellCountsPerCluster.SingleCellExperiment, for example.
• Cleaned up example datasets. Removed known_markers_detected_small and sce_small.
• diffExp: Reorganized and reworked internal code that uses zinbwave.
• Reworked and improved Seurat-to-SingleCell interconversion support. Refer to seurat-SingleCellExperiment-methods for details.

Deprecations:

These functions have been deprecated: knownMarkers, knownMarkersDetected, plotKnownMarkersDetected, readCellTypeMarkers, sanitizeMarkers.

pointilism 0.1.3 (2018-09-19)

Major changes:

• as(seurat, "SingleCellExperiment") now returns scaleData in assays, along with varGenes in metadata and ensures that colData and rowRanges are sanitized into camel case using the camel function internally.

Minor changes:

• Removed convertGenesToSymbols support for seurat class. Coerce genes to symbols manually prior to seurat coercion from SingleCellExperiment.
• clusterID: improved consistency between SingleCellExperiment and seurat methods, which now both work on colData.
• Improved internal fetchData calls to use camel case for consistency.
• plotFeature now sanitizes the features argument into camel case internally, for more consistent matching.
• plotReducedDim now uses camel case formatting for axis labels.
• Removed unnecessary legacy .minimalAxis constructor.
• metadata<- assignment method now stashes directly into object@misc for seurat class.
• Updated example data script and resaved.

pointillism 0.1.2 (2018-08-22)

New functions:

• clusterID: S4 generic that returns the cell cluster identity mappings as a factor for all cells.
• findMarkers: Utility function that wraps diffExp to identify cluster- specific genes across all clusters.

Minor changes:

• Converted runZinbwave to an S4 method that works on SingleCellExperiment. For seurat objects, coerce to SingleCellExperiment before calculating weights.
• plotDot now uses default ggplot2 color scale, unless specified.
• plotReducedDim, plotMarker: theme_paperwhite is no longer used by default. Removed grid and aspectRatio arguments to simplify ggplot2 code handling.

pointillism 0.1.1 (2018-08-20)

New functions:

• Now exporting runZinbwave and hasZinbwave, which were previously internal functions called by diffExp.

Minor changes:

• Improved internal handling of assays for objects that have zinbwave calculations applied. Now it doesn't attempt to keep only the counts assay.

pointillism 0.1.0 (2018-08-10)

• Initial release.

steinbaugh/pointillism documentation built on Oct. 13, 2023, 10:43 p.m.