NEWS.md
output_width and output_height to the Iterate_* family of plotting functions (#217).Random_Cells_Downsample that prevented setting identity using the group.by parameter.Cell_Highlight_Plot that didn’t pass the reduction parameter properly (#216).return_plots in iterative plotting functions(#217).Major Updates to Functionality with rliger Package:
Added new utility functions to interact with liger v2.0.0+ object format change:
- Subset_LIGER to quickly subset by cluster or other meta data variable.
- Cells_by_Identities_LIGER to extract list of barcodes sorted by values within given meta data column.
Extended the following Seurat/SeuratObject generic functions to work seamlessly with liger objects:
- Cells to extract vector of all cells or list vectors of cells by dataset.
- Features to extract vector of all features or list vectors of features by dataset.
- WhichCells to extract vector or list of cells matching identity criteria.
- Embeddings to extract matrix containing dimensionality reduction embeddings or iNMF h.norm matrix.
- Idents and Idents<- to extract and set default identities/clusters.
Updated functions to interact with both old and new style liger objects:
- plotFactors_scCustom()
- Fetch_Meta
- Top_Genes_Factor
- Add_Mito_Ribo
- Add_Cell_Complexity
- DimPlot_LIGER
- Variable_Features_ALL_LIGER
- Feature_Present
New functions compatible with old and new style liger objects:
- Added new function Add_Hemo to add hemoglobin gene percentage for QC. Also added as parameter to Add_Cell_QC_Metrics. Add_Hemo supports all default species: (human, mouse, marmoset, zebrafish, rat, drosophila, rhesus macaque, and chicken) and works with both Seurat and liger objects.
New scCustomize generics to function across both Seurat and Liger objects:
- Add_Hemo (see above).
- Rename_Clusters now S3 generic for setting new active.ident (Seurat) or defaultCluster (Liger).
New functions for Seurat and rliger v2.0.0+ only:
- Added new function Find_Factor_Cor to return correlation matrix between factor gene loadings from liger or Seurat object.
- Added new function Factor_Cor_Plot to plot positive correlations from liger or Seurat object.
Updated functions to recommend new rliger equivalents for users with rliger v2.0.0+:
- as.LIGER
- as.Seurat
General scCustomize Updates:
New functions:
- Added new function Add_Hemo to add hemoglobin gene percentage for QC. Also added as parameter to Add_Cell_QC_Metrics. Add_Hemo supports all default species: (human, mouse, marmoset, zebrafish, rat, drosophila, and rhesus macaque) and works with both Seurat and liger objects.
- Added new function seq_zeros() to create sequences with preceding zeros.
- Added new function Read_Metrics_CellBender to read in the summary metrics csv file produced by CellBender. Can either read all metrics files from parent directory of output folders or a single metrics file.
- Added Updated_MGI_Symbols to check for update gene names/symbols in mouse data (#202).
- Added plotting function Proportion_Plot to plot pie chart or bar chart of proportion (or total counts) of cells in each identity class.
- Added new function Random_Cells_Downsample to return either a vector or list with randomly downsampled cells for each identity class.
- Added new function Cells_per_Sample to quickly return data.frame with just number of cells per sample.
Updated functions:
- Added new parameters data_name and overwrite to Add_Alt_Feature_ID to support new storage location.
- Added cells parameter explicitly to FeatureScatter_scCustom.
- Added Chicken (Gallus gallus) to default species for QC functions. Thanks @dpearton; (#176).
- Added new plotting function SpatialDimPlot_scCustom. Thanks for encouragement @puapinyoying @nina-hahn (#160).
- Added ability of Read_Metrics_10X to read a single metrics csv file and return data formatted the same way as when reading multiple files.
- Added parameter cutoff_line_width to the QC_Plot_* family of plots to control line thickness of cutoff lines.
- Cluster_Stats_All_Samples now returns data.frame with row order reflecting the frequency of cells.
- Add_Mito_Ribo now supports datasets aligned to multi-species reference genomes (#184).
- Added parameter add_prop_plot to DimPlot_scCustom to return plot showing number or percent of cells per identity along with the DimPlot.
- Added optional parameter colors_use_assay2 to FeaturePlot_DualAssay which allows for specification of different palettes for the two plots (#182).
- Added new folder and scripts (see “data-raw/” on GitHub) detailing the creation of gene lists used in Add_Cell_QC_Metrics.
- Added ensembl ID support for percent hemoglobin, msigdb, and IEG gene sets (#186).
- Add verbosity parameter to Store_Misc_Info_Seurat and Store_Palette_Seurat.
- Explicitly reveal the reduction parameter in Cluster_Highlight_Plot and Meta_Highlight_Plot (#198).
- Added show_row_names show_column_names, column_names_side, row_names_side, legend_position, legend_orientation, show_ident_legend, and show_ident_colors parameters to Clustered_DotPlot. Thanks for idea and code @johnminglu (#199).
- Updated Split_Vector to allow user to specify number of chunks or size of chunks for splitting vector.
- Update RenameClusters with additional parameters to enable storage of both old idents and new idents in meta.data within the function.
- Update Add_Cell_QC_Metrics.Seurat to explicitly reveal list_species_names parameter.
- Added new vignette for spatial plotting.
- Added new and expanded vignette on use of object QC functions for better clarity on these functions and their uses (previously was part of QC Plotting & Helpers/Utilities Vignettes). Plotting elements of QC Plotting vignette are unchanged.
Add_Top_Gene_Pct_Seurat is now S3 generic that works with both Seurat and liger objects and has been renamed Add_Top_Gene_Pct.Add_Cell_QC_Metrics is now S3 generic and works with both Seurat and liger objects.Add_Alt_Feature_ID to @misc slot of object for safer storage across object filtering.as.anndata to explicitly check for installation of anndata before starting conversion (#162).Plot_Median_Genes, Plot_Median_UMIs, Plot_Median_Mito, Plot_Median_Other, Plot_Cells_per_Sample to understand “ident” as grouping variable.Store_Misc_Info_Seurat to use Seurat accessor/setter function Seurat::Misc().sample_names in Read_CellBender_h5_Multi_File to clarify parameter behavior (related to (#208)).Read_Metrics_10X to support adjusts to metrics summary format and metric names in output from Cell Ranger v9+.Plot_Density_Custom and Plot_Density_Joint_Only have been re-enabled for users with ggplot2 v3.5.0 following Nebulosa v1.12.1 update patch.Add_Cell_QC_Metrics when overwrite = TRUE (#165).DotPlot_scCustom (#158).as.anndata from Seurat conversion that previously caused failures (#168).Create_Cluster_Annotation_File if for file path and csv name errors.plot_median and more than one feature in VlnPlot_scCustom (#169).DimPlot_scCustom due to changes in guides updated with ggplot2 v3.5.0 (#171).Add_Sample_Meta that still errored when setting na_ok = TRUE.Plot_Median_* family that caused issues when group_by parameter was NULL.FeaturePlot_scCustom when setting combine = FALSE.DimPlot_scCustom that could cause blank plot when rasterizing points.MAD_Stats that didn’t respect mad_num parameter (#183).MAD_Stats that could cause issues if mad_num was less than or equal to 0 and returned error if setting group_by_var to “ident”.all_of/any_of syntax.Cluster_Highlight_Plot with split.by parameter (#201).Convert_Assay (#205).Updated_HGNC_Symbols due to change in URL path for gene names (#209).DimPlot_scCustom when split.by and label.box = TRUE.DiscretePalette_scCustom that didn’t error when supplying invalid palette names.DimPlot_LIGER that provided uniformative error message when changing the default cluster ident.title_prefix parameter to Iterate_DimPlot_bySample to unify with Meta_Highlight_Plot.Split_Vector to split vector in chunks of predetermined sizes.Updated_HGNC_Symbols to update human gene symbols. After first use does not require internet connection.Seurat::LogSeuratCommand().plot_legend to plotFactors_scCustom to allow for removal to legend in factor loading plots.Iterate_FeaturePlot_scCustom to allow for plotting multiple plots per page when saving to single PDF document (see new parameters: features_per_page and landscape.LIGER_Features utility function for LIGER objects (analogous to Seurat::Features()).as.LIGER() as enhanced method for conversion of Seurat objects or lists of Seurat objects to single LIGER object.as.anndata() to support conversion of Seurat and LIGER objects to anndata format.Convert_Assay() to perform easy conversions of Seurat assays between V3/4 (Assay) and V5 (Assay5) formats.assay_type to allow manual control of assay type when creating Seurat object from LIGER using Liger_to_Seurat. Now as.Seurat.grid_color to Clustered_DotPlot to control the color of grid lines (default is no grid lines).Clustered_DotPlot by additional variable and maintain expression information.Split_Layers() function for V5 objects.Add_Alt_Feature_ID to add alternative feature ids to an Assay5 meta.data slot.Add_Mito_Ribo() replaces Add_Mito_Ribo_Seurat and Add_Mito_Ribo_LIGER.Add_Cell_Complexity() replaces Add_Cell_Complexity_Seurat and Add_Cell_Complexity_LIGER.Meta_Present_LIGER has been deprecated and wrapped inside of Meta_Present.Liger_to_Seurat() has been soft-deprecated. It is replaced by new extension of Seurat generic as.Seurat with added support for Liger objects, using all the same parameters as Liger_to_Seurat. Full deprecation will occur in v2.2.0.Gene_Present has been soft-deprecated. It is replaced by Feature_Present which functions identically but better reflects that features present may also be proteins. Full deprecation will occur in v2.2.0.legend in Iterate_DimPlot_bySample has been inverted to no_legend to match Meta_Highlight_Plot parameters.Liger_to_Seurat() for compatibility with Seurat V5 structure (#153). Now part of as.Seurat.Stacked_VlnPlot due to rasterization defaults.alpha_na_exp appropriately in FeaturePlot_scCustom.alpha_exp between Seurat versions 4 and 5 (#144).DimPlot_scCustom when levels are missing from a split plot.FeaturePlot_scCustom that could cause plots to be mislabeled when using split.by and depending on the order of features provided (#150).DimPlot_LIGER (#153).Add_Mito_Ribo_LIGER that caused it to return value of 0 for all cells (Now part of renamed Add_Mito_Ribo S3 generic).Clustered_DotPlot to display percentage instead of proportion to match legend text.Percent_Expressing error when group_by = "ident".DotPlot_scCustom that didn’t correctly pass group.by when plotting (#158).Add_Mito_Ribo_Seurat causing failure due to error message when overwrite = TRUE.Add_Top_Gene_Pct_Seurat to avoid issue that accidentally could call function on normalized data.Add_Top_Gene_Pct_Seurat that caused error if more than one counts layer was present.QC_Histogram that prevented plotting or titling of plots.multi pipeline to Read10X_Metrics via new parameter cellranger_multi.dot_size parameter to Seq_QC_Plot_* family of functions.Barcode_Plot and Iterate_Barcode_Rank_Plot.ident_legend parameter to QC_Plot_UMIvsGene to control show/hide of the identity legend (#121).CellBender_Feature_Diff.min_count_label in CellBender_Diff_Plot to better control feature labeling.Iterate_DimPlot_bySample using new sample_column parameter.MAD_Stats to calculate to the median absolute deviation of meta.data columns by grouping variable and across entire object.Add_Top_Gene_Pct_Seurat to add another QC measure of cell complexity to object meta.data. Returns percentage of counts occupied by top XX genes in each cell.VariableFeaturePlot_scCustom using custom_features parameter.Add_Cell_QC_Metrics to simplify adding cell QC metrics. Single function call to add Mito/Ribo Percentages, Cell Complexity, Top Gene Percentages, MSigDB Percentages, IEG Percentages, and/or Cell Cycle Scoring (human only).Add_Cell_QC_Metrics function: “msigdb_qc_gene_list” and “ieg_gene_list”.Add_Cell_QC_Metrics.plot_median and plot_boxplot to VlnPlot_scCustom (and VlnPlot_scCustom-based plots; e.g., QC_Plot_* family) for added visualization.QC_Histogram to plot QC features (or any feature) using simple histogram.FeatureScatter_scCustom function to customize Seurat’s FeatureScatter plots.figure_plot parameter to all 2D DR (t-SNE, UMAP, etc) based plots (#127).Split_FeatureScatter function has been completely deprecated and it’s functionality has been moved to new FeatureScatter_scCustom.gene_list in Iterate_FeaturePlot_scCustom and Iterate_VlnPlot_scCustom has been soft-deprecated and replaced by features parameter. Specifying gene_list will display deprecation warning but continue to function until next major update.Read10X_Metrics to use new internal helper functions.Liger_to_Seurat to transfer the liger_object@H slot in addition to H.norm slot already moved.length(x = colnames(x = obj) with length(x = Cells(x = obj) for accurate plotting based on V5 object structure.Gene_Present now accepts assay parameter.R/ for better organization.scCustomize_Palette). Now if number of colors is greater than 2 but less than 8 the default palette will be ColorBlind_Pal (previously it was “polychrome”). Polychrome remains the default when number of colors is between 9-36.scCustomize_Palette to ggplot_default_colors = FALSE to avoid uncessary error when no value supplied.DimPlot_scCustom will now set label = TRUE if label.box is set to TRUE but label is not changed from default.PackageCheck (now deprecated), with rlang::is_installed() for non-dependency checks.Read10X_Metrics that caused errors when reading files on windows operating system (#115).Create_CellBender_Merged_Seurat when feature names are changed (underscore to dash) during object creation (#118).Read10X_h5_Mutli_Directory when reading Cell Ranger multi directories.VlnPlot_scCustom, DimPlot_scCustom, and DotPlot_scCustom to avoid otherwise ambiguous error messages (#120).VlnPlot_scCustom (#122).Cell_Highlight_Plot that could cause function to error without proper error message.Read_* functions to avoid unnecessary errors.drop_na(.data[[var]], with current dplyr syntax.VlnPlot-based plots when split.by is not NULL.group.by when using DimPlot_scCustom (#128).Add_Mito_Ribo_Seurat and Add_Mito_Ribo_LIGER which incorrectly stated the names of new meta.data/cell.data columns to be added.DotPlot_scCustom that prevented it from working unless group.by parameter was explicitly added.Case_Check caused by typo.Cluster_Highlight_Plot and Meta_Highlight_Plot that were too verbose.Add_Mito_Ribo_Seurat and Add_Mito_Ribo_LIGER which caused error when supplying custom list of features for non-default organism (#133).DimPlot_scCustom preventing that errored when trying to split plot and use figure_plot at same time.aspect_ratio parameter to all dimensionality reduction plots to control axes ratio of output plot.plot_median and median_size parameters to QC_Plots_* functions.split_collect parameter to FeaturePlot_scCustom to collect all guides when using split.by for a single feature (#94).Clustered_DotPlot to allow modification of sizes of column text labels, legend text labels, and legend title labels (#96).Merge_Sparse_Multimodal_All for merging multi-modal data (1 matrix per modality) (#104).Clustered_DotPlot named row_label_fontface to allow control of fontface used for row labels (#103).Reduction_Loading_Present, in part to fix issue with FeaturePlot_scCustom and internal feature checking.Clustered_DotPlot using new parameters: cluster_feature, cluster_ident (#106).dot_size parameter to statistics plotting functions Plot_Cells_per_Sample and Plot_Median_* family.no_legend to Iterate_Meta_Highlight_Plot to allow for plotting with a plot title instead of plot legend (#108).QC_Plots_Feature to use VlnPlot_scCustom under the hood like rest of QC_Plots_* functions.abort in Meta_Present to return_none to align with Gene_Present and Reduction_Loading_Present.summarise_at, select(.data[[var]]), and rename(.data[[var]]) with current dplyr syntax.Iterate_Cluster_Highlight_Plot and Iterate_Meta_Highlight_Plot to align with recent updates to base Cluster_Highlight_Plot and Meta_Highlight_Plot functions.QC_Plots_Feature to respect parameters when passing to VlnPlot (#91).Read_CellBender_h5_* functions to support CellBender outputs from STARsolo- or Cell Ranger (pre-V3)-processed data (#99).FeaturePlot_scCustom to allow for plotting of dimensionality reduction loadings (#97).Read10X_Multi_Directory and Read10X_h5_Multi_Directory to support files processed with Cell Ranger multi pipeline.Merge_Seurat_List that prevented add.cell.id from adding correct cell name prefixes (#113).label_color_num parameter to PalettePlot allow control of color labeling.Stacked_VlnPlot 90 degrees or 45 (previously possible) (#84).Merge_Seurat_List to avoid ambiguous error messages on failure.Case_Check checks/messages to all feature-based plotting functions.PalettePlot has been changed from palette to pal.PalettePlot to support pal of class “colors”.Cluster_Highlight_Plot and Meta_Highlight_Plot.Fetch_Meta is now S3 generic function that can handle either Seurat or LIGER objects.R/ scripts for better organization.merge parameter to Read10X_GEO, Read10X_h5_GEO, Read_GEO_Delim and Read_CellBender_h5_Multi_File.raster.dpi parameter to DimPlot_LIGER.label parameter to FeaturePlot_scCustom to avoid error collision (#80).vln_linewidth parameter to control violin outline line width (#32).Fetch_Meta for returning data.frame of object meta data.Extract_Sample_Meta to extract sample-level meta data from object.Cell_Highlight_Plot for highlight plots of custom cells not in active ident or meta data.flip parameter to Clustered_DotPlot to enable axes flipping (#69).CellBender_Feature_Diff to return data.frame with count sums and differences between raw and CellBender assays.CellBender_Diff_Plot to plot differences between raw and CellBender assays using data from CellBender_Feature_Diff.mito_name parameter to QC_Plots_Mito to allow for custom specification of meta data column name that contains mitochondrial information.QC_Plots_Combined_Vln() function to return patchwork layout of 3 QC plots.Add_Mito_Ribo_Seurat() and Add_Mito_Ribo_LIGER() (#28).alpha_exp and alpha_na_exp parameters to FeaturePlot_scCustom to allow for control of color scale transparency (#21).*_Highlight_Plot functions can now plot multiple variables simultaneously using either one color for all variables or one color per variable (#34).figure_plot to DimPlot_scCustom(). This removes axes and axes labels and adds axis legend on left bottom corner of plot (#40).plot_legend to Stacked_VlnPlot. This solves issue with returning only one shared legend across all features being plotted (#48).Add_Cell_Complexity_Seurat and Add_Cell_Complexity_LIGER functions to add cell QC complexity/novelty metric (log10(Genes) / log10(UMIs)).QC_Plots_Complexity plot for quick plotting of cell complexity score.Read_CellBender_h5_Mat, Read_CellBender_h5_Multi_Directory, Read_CellBender_h5_Multi_File to enable easy reading of new CellBender output files.raster.dpi parameter from Seurat to all DimPlot FeaturePlot or FeatureScatter based functions.add.noise parameter from Seurat to VlnPlot_scCustom Stacked_VlnPlot functions.group.by as default listed parameter to added to allVlnPlot based QC_Plot_*.ensembl_ids parameter for Add_Mito_Ribo_* functions. If ensembl_ids = TRUE functions will retrieve stored ensembl IDs representing mitochondrial and ribosomal genes for accepted default species.label_feature_yaxis to FeaturePlot_scCustom. Allows for plotting of feature names on secondary y-axis when using split.by (#60).Add_Sample_Meta function for addition of sample-level meta data to cell-level @meta.data slot of Seurat objects.Read_GEO_Delim to check for issues with imported matrices. Check is modified version of SeuratObject::CheckMatrix called CheckMatrix_scCustom(). Will warn if infinite, logical, non-integer (whole), or NA/NaN values are detected in input matrix.QC_Plot_UMIvsGene will now returned filtered correlation value that takes into account meta_gradient_name if provided in addition to nFeature_RNA and nCount_RNA.Variable_Features_ALL_LIGER which allows for detection/selection of variable genes from entire LIGER object instead of iterating by dataset.PackageCheck warning has been added where needed.VlnPlot has been changed to Iterate_VlnPlot_scCustom to reflect that it now uses VlnPlot_scCustom to generate plots.QC_Plot_* functions now use VlnPlot_scCustom internally to unify color scheme and rasterization parameters.*_Highlight_Plot functions no longer display “Unselected” in plot legend and uses DimPlot_scCustom to generate plots (#34).low_cutoff from plotting via QC_Plots_Mito.Clustered_DotPlot that prevented setting identity colors (#29).FeaturePlot_scCustom that returned NULL when setting combine = FALSE (#31).Seq_QC_Plot_* functions which resulted in groups being plotted out of order when specifying plot_by parameter.Seq_QC_Plot_* functions that created color palette error when color palettes were not being used.DimPlot_scCustom that caused mismatch of colors between plots when using split.by if one of the plots was missing 1 or more of the group.by levels (#37).VlnPlot_scCustom that caused raster warning messages to be displayed twice (#42).Iterate_PC_Loading_Plots that caused error when specifying current directory with file_path = NULL or file_path = ""
DotPlot_scCustom that prevented plotting of features in meta.data slot (#44).Stacked_VlnPlot when no supplied features were present.plotFactors_scCustom that was ignoring provided file name.plotFactors_scCustom that caused progress to only display progress up to 50% even when it was fully complete.Clustered_DotPlot that resulted in error related to color palettes if number of clusters was greater than 36 (#49).Add_Mito_Ribo_LIGER that resulted custom column names (e.g. mito_name = "pct.mt") being disregarded and also therefore issue with overwrite parameter. (#51).Store_Misc_Info_Seurat that prevented function from working.Plot_Density_Custom when supplying custom_palette and multiple features. (#51).Clustered_DotPlot so that legend with identities is displayed by factor level of Seurat object idents (#55).Split_FeatureScatter to remove test code that prevented function from working properly (#57).DimPlot_All_Samples, Split_FeatureScatter, and DimPlot_scCustom that ignored factor order when plotting groups.FeaturePlot_scCustom when setting split.by and one or more of features provided was not present in object (#64).VlnPlot_scCustom function.Stacked_VlnPlot
make_unique parameter to Extract_Top_Markers function.Clustered_DotPlot function.Add_Mito_Ribo_Seurat and Add_Mito_Ribo_LIGER.Merge_Sparse_Data_All. (#8)Read_Metrics_10X, Read10X_Multi_Directory, and Read10X_h5_Multi_Directory.StackedVlnPlot with parameters plot_spacing and spacing_unit. (#6)scCustomize_Palette function select palette to use (simplify internal code).