vertesy / seurat.utils

Various utility functions for Seurat single-cell analysis

Seurat.utils

Seurat.utils Is a collection of utility functions for Seurat. Functions allow the automation / multiplexing of plotting, 3D plotting, visualisation of statistics & QC, interaction with the Seurat object, etc. Some functionalities require functions from CodeAndRoll2, ReadWriter, Stringendo, ggExpressDev, MarkdownReports, and the Rocinante (See installation).

[TOC]

Installation

Seurat.utils relies on:

• Stringendo
• ReadWriter
• CodeAndRoll2
• MarkdownHelpers
• Markdownreports
• ggExpress

... and provides functions for

• Seurat.pipeline

You can install all of them directly from GitHub via devtools with one R command:

# install.packages("devtools"); # If you don't have it.
BiocManager::install("sparseMatrixStats")
require("devtools")

# Install dependencies
devtools::install_github(repo = "vertesy/Stringendo", upgrade = F)
devtools::install_github(repo = "vertesy/CodeAndRoll2", upgrade = F)
devtools::install_github(repo = "vertesy/ReadWriter", upgrade = F)
devtools::install_github(repo = "vertesy/MarkdownHelpers", upgrade = F)
devtools::install_github(repo = "vertesy/Markdownreports", upgrade = F)
devtools::install_github(repo = "vertesy/ggExpress", upgrade = F)

# Recommended
devtools::install_github(repo = "vertesy/DatabaseLinke.R", upgrade = F)

# Install Seurat.utils
devtools::install_github(repo = "vertesy/Seurat.utils", upgrade = F)

# You may need to install
# install.packages("pheatmap")
# install.packages("checkmate")

...then simply load the package:

require("Seurat.utils")

**NOTE: If you type 'all' when R asks to update dependencies, you may get into installation errors / infinite loops. If updating fails, type 'no' when prompted. **

Alternatively, you simply source it from the web. This way function help will not work, and you will have no local copy of the code on your hard drive.

source("https://raw.githubusercontent.com/vertesy/Seurat.utils/master/R/Seurat.Utils.R")

Troubleshooting

If you encounter a bug, something doesn't work or unclear, please let me know by raising an issue on Seurat.utils – Please check if it has been asked.

Error on install

• Dependencies should be installed in the specific order.
• Error: Failed to install 'unknown package' from GitHub: HTTP error 403. API rate limit exceeded -> Try connecting later or via another wifi
• Try updating all packages update.packages()

Error during usage

#1 Check and reinstall dependencies

If you have an older installation, it is quite possible that some of the above packages are out of date. Please reinstall all of them, in order (see: Installation).

Some functionalities are coming from dependencies. If the error points to a function outside or Seurat.utils, check in the packages below:

• CodeAndRoll2
• ReadWriter
• Stringendo
• MarkdownReports

#2 Maybe you need Rocinante.

It needs to be sourced, cannot be installed:

source("https://raw.githubusercontent.com/vertesy/Rocinante/master/R/Rocinante.R")

List of Functions

List of Functions in Seurat.Utils.R (94)

Updated: 2024/03/22 15:41

• 1 parallel.computing.by.future()

  parallel.computing.by.future. Run gc(), load multi-session computing and extend memory limits.

• 2 IntersectGeneLsWithObject()

  Intersect Genes with Seurat Object. Intersects a set of gene names with those found in a Seurat object.

• 3 SelectHighlyExpressedGenesq99()

  Intersect Genes with the List of Noticeably Expressed Genes. Intersects a vector of gene names with a Seurat object to find genes that are both in the input list and have expression levels in the top quantiles as defined by the object's q99 expression data. It aims to filter genes based on their expression levels being above a specified threshold. Additionally, it offers an option to sort the genes by their expression levels in decreasing order.

• 4 SmallestNonAboveX()

  SmallestNonAboveX. replace small values with the next smallest value found, which is >X.

• 5 Create.MiscSlot()

  AreTheseCellNamesTheSame. Assert and compare two character vectors (e.g.: cell IDs) how much they overlap and plot a Venn Diagram. The function aborts with an error if overlap is too small.

• 6 addToMiscOrToolsSlot()

  Add to Misc or Tools Slot. This function creates and adds a sub-slot to either the 'misc' or 'tools' slot of a Seurat object. If the sub-slot already exists, it can either be overwritten or a warning will be issued.

• 7 showToolsSlots()

  Display Slots in the @tools of an Seurat Object. showToolsSlots prints the names of slots in the @tools of a given object. It specifically targets list elements, skipping over data frames and other non-list objects.

• 8 showMiscSlots()

  Display Slots in the @misc of an Seurat Object. See showToolsSlots for details. Prints the names of slots in the @misc of a given object. It specifically targets list elements, skipping over data frames and other non-list objects.

• 9 calc.q99.Expression.and.set.all.genes()

  calc.q99.Expression.and.set.all.genes. Calculate the gene expression of the e.g.: 99th quantile (expression in the top 1% cells).

• 10 RenameClustering()

  RenameClustering. Rename clustering in a Seurat object.

• 11 getClusterNames()

  Shorten Clustering Names. This function takes in a string representing a clustering name, and shortens it according to specific rules. It replaces "snn_res." with "", "best.matching.names" with "bmatch", "ordered" with "ord", "ManualNames" with "mNames", and ".long" at the end of the string with ".L".

• 12 GetClusteringRuns()

  GetClusteringRuns. Get Clustering Runs: metadata column names.

• 13 GetNamedClusteringRuns()

  GetNamedClusteringRuns. Get Clustering Runs: metadata column names

• 14 GetOrderedClusteringRuns()

  GetOrderedClusteringRuns. Get Clustering Runs: metadata column names.

• 15 GetNumberOfClusters()

  GetNumberOfClusters. Get Number Of Clusters #

• 16 calc.cluster.averages()

  calc.cluster.averages. Calculates the average of a metadata column (numeric) per cluster.

• 17 plot.expression.rank.q90()

  plot.expression.rank.q90. Plot gene expression based on the expression at the 90th quantile (so you will not lose genes expressed in few cells).

• 18 BackupReduction()

  BackupReduction. Backup UMAP to obj@misc$reductions.backup from obj@reductions$umap. #

• 19 SetupReductionsNtoKdimensions()

  SetupReductionsNtoKdimensions. Function to calculate N-to-K dimensional umaps (default = 2:3); and back them up to slots obj@misc$reductions.backup from @reductions$umap

• 20 RecallReduction()

  RecallReduction. Set active UMAP to obj@reductions$umap from obj@misc$reductions.backup.

• 21 set.mm()

  set.mm. Helps to find metadata columns. It creates a list with the names of of '[email protected]'.

• 22 ww.get.1st.Seur.element()

  Get the First Seurat Object from a List of Seurat Objects. #' If provided with a list of Seurat objects, this function returns the first Seurat object in the list. If the input is a single Seurat object, it returns the object itself. It is assumed that all elements of the list are Seurat objects if the input is a list.

• 23 recallAllGenes()

  recallAllGenes. all.genes set by calc.q99.Expression.and.set.all.genes() #

• 24 recall.meta.tags.n.datasets()

  recall.meta.tags.n.datasets. Recall meta.tags from obj@misc to "meta.tags" in the global environment.

• 25 recall.parameters()

  recall.parameters. Recall parameters from obj@misc to "p" in the global environment.

• 26 recall.genes.ls()

  recall.genes.ls. Recall genes.ls from obj@misc to "genes.ls" in the global environment.

• 27 save.parameters()

  Save Parameters to Seurat Object. Stores a list of parameters within the @misc$p slot of a Seurat object, allowing for easy reference and tracking of analysis parameters used.

• 28 create_scCombinedMeta()

  Create Single-Cell Metadata Object for a collection of Seurat Objects. This function creates a metadata object to correspond to a list of single-cell experiments, for storing parent level information. It initializes the object with the experiment and project name, and the creation date. The created object is of class 'scMetadata_class'.

• 29 copyMiscElements()

  Copy Specified Elements from One Seurat Object's @misc to Another's. Copies specified elements from the @misc slot of one Seurat object to the @misc slot of another. It warns if some specified elements are missing in the source object or if elements are overwritten in the destination object, depending on the overwrite argument.

• 30 copyCompleteToolsSlots()

  Copy Tools Slots from Multiple Seurat Objects. This function copies the @tools slots from a list of Seurat objects into a new slot of a target Seurat object. This allows for the aggregation of tools information from multiple experiments or datasets into a single, consolidated Seurat object.

• 31 subsetSeuObjByIdent()

  Subset a Seurat Object by Identity. Subsets a Seurat object based on a specified identity column and values. It allows for an optional inversion of the selection.

• 32 downsampleSeuObj()

  downsampleSeuObj. Subset a compressed Seurat object and save it in the working directory.

• 33 downsampleSeuObj.and.Save()

  downsampleSeuObj.and.Save. Subset a compressed Seurat Obj and save it in wd. #

• 34 downsampleSeuObjByIdentAndMaxcells()

  Sample Cells From Identifiers in Seurat Object. This function samples a specified maximum number of cells from each identity class in a Seurat object, in the meta.data. It ensures that the sampling does not exceed the total number of cells available per identity.

• 35 removeResidualSmallClusters()

  Remove Residual Small Clusters from Seurat Object. Removes clusters containing fewer cells than specified by max.cells from a Seurat object. This function is particularly useful after subsetting a dataset, where small, possibly unrepresentative clusters may remain.

• 36 dropLevelsSeurat()

  dropLevelsSeurat. Drop unused levels from factor variables in a Seurat object.

• 37 removeClustersAndDropLevels()

  Remove Clusters and Drop Levels. This function removes residual small clusters from specified Seurat objects and drops levels in factor-like metadata.

• 38 removeCellsByUmap()

  Remove Cells by Dimension Reduction. This function applies a cutoff in the specified dimension of a given dimension reduction (UMAP, PCA, or t-SNE) to remove cells.

• 39 downsampleListSeuObjsNCells()

  Downsample a List of Seurat Objects to a Specific Number of Cells. Downsampling each Seurat object in a list to a specified number of cells. This function is particularly useful for creating smaller, more manageable subsets of large single-cell datasets for preliminary analyses or testing.

• 40 downsampleListSeuObjsPercent()

  Downsample a List of Seurat Objects to a Fraction. Downsampling a list of Seurat objects to a specified fraction of their original size. This is useful for reducing dataset size for quicker processing or testing workflows.

• 41 Add.DE.combined.score()

  Add.DE.combined.score. Add a combined score to differential expression (DE) results. The score is calculated as log-fold change (LFC) times negative logarithm of scaled p-value (LFC * -log10( p_cutoff / pval_scaling )).

• 42 StoreTop25Markers()

  Save Top 25 Markers per Cluster. Stores the top 25 markers for each cluster identified in a Seurat object, based on the avg_log2FC from the output table of FindAllMarkers(). The result is saved under @misc$df.markers$res..., rounding insignificant digits to three decimal places.

• 43 StoreAllMarkers()

  Store All Differential Expression Markers. Saves the complete output table from FindAllMarkers() to a Seurat object, facilitating easy access to differential expression analysis results. This function rounds numerical values to a specified number of digits to maintain readability and manage file sizes.

• 44 GetTopMarkersDF()

  Get Top Differential Expression Genes Data Frame. Retrieves a data frame of the top N differentially expressed genes from differential gene expression analysis results, offering an option to exclude certain genes based on patterns.

• 45 GetTopMarkers()

  Get Top Differential Expression Markers from DGEA Results. Retrieves the top N differentially expressed genes from the results of a differential gene expression analysis, such as that provided by FindAllMarkers().

• 46 AutoLabelTop.logFC()

  AutoLabelTop.logFC. Create a new "named identity" column in the metadata of a Seurat object, with Ident set to a clustering output matching the res parameter of the function. It requires the output table of FindAllMarkers(). If you used StoreAllMarkers() is stored under @misc$df.markers$res..., which location is assumed by default.

• 47 AutoLabel.KnownMarkers()

  AutoLabel.KnownMarkers. Creates a new "named identity" column in the metadata of a Seurat object, setting 'Ident' to a clustering output matching the 'res' parameter. This function requires the output table of FindAllMarkers(). If you used StoreAllMarkers(), the output is stored under @misc$df.markers$res..., which is the default location.

• 48 # sparse.cor4()

  Calculate Sparse Correlation Matrix. Computes a sparse correlation matrix from a given sparse matrix input. This function is useful for efficiently handling large datasets where most values are zero, facilitating the calculation of both covariance and correlation matrices without converting to a dense format.

• 49 Calc.Cor.Seurat()

  Calc.Cor.Seurat. Calculate gene correlation on a Seurat object.

• 50 plot.Gene.Cor.Heatmap()

  Plot Gene Correlation Heatmap. Generates a heatmap visualization of gene correlations based on expression data. Useful for identifying groups of genes that exhibit similar expression patterns across different conditions or cell types in a Seurat object.

• 51 prefix_cells_seurat()

  Add Prefixes to Cell Names in Seurat Objects. Adds prefixes derived from a vector of identifiers to cell names in a list of Seurat objects. This is useful for ensuring unique cell names across multiple samples or conditions when combining or comparing datasets.

• 52 find_prefix_in_cell_IDs()

  Check Prefix in Seurat Object Cell IDs. This function checks if a prefix has been added to the standard cell-IDs (16 characters of A,TRUE,C,G) in a Seurat object. If so, it prints the number of unique prefixes found, issues a warning if more than one unique prefix is found, and returns the identified prefix(es).

• 53 seu.Make.Cl.Label.per.cell()

  Create Cluster Labels for Each Cell. Generates labels for each cell by combining gene names and cluster IDs. This function takes a named vector, typically representing top genes for clusters (values) and their corresponding cluster IDs (names), along with a vector of cell IDs. It then creates a new vector where each cell is labeled with its top gene and cluster ID in the format "GeneName.ClusterID".

• 54 GetMostVarGenes()

  Retrieve the Top Variable Genes from a Seurat Object. Retrieves the names of the most variable genes from a Seurat object, typically used to focus subsequent analyses on genes with the greatest variation across cells.

• 55 gene.name.check()

  Check Gene Names in Seurat Object. Examines gene names in a Seurat object for specific naming conventions, such as the presence of hyphens (-) or dots (.) often found in mitochondrial gene names. This function is useful for ensuring gene names conform to expected patterns, especially when preparing data for compatibility with other tools or databases.

• 56 check.genes()

  Check if Gene Names exist in Seurat Object or HGNC Database. Verifies the presence of specified gene names within a Seurat object or queries them against the HGNC database. This function is useful for ensuring gene names are correctly formatted and exist within the dataset or are recognized gene symbols.

• 57 fixZeroIndexing.seurat()

  Fix Zero Indexing in Seurat Clustering. Adjusts Seurat object metadata to fix zero-based cluster indexing, converting it to one-based indexing. This function modifies a specified metadata column in the Seurat object to replace zero-indexed cluster names with one-based indexing.

• 58 CalculateFractionInTrome()

  Calculate Fraction of Genes in Transcriptome. Calculates the fraction of specified genes within the entire transcriptome of each cell in a Seurat object. This function is useful for assessing the relative abundance of a set of genes across cells, such as identifying cells with high expression of marker genes.

• 59 AddNewAnnotation()

  AddNewAnnotation. This function creates a new metadata column based on an existing metadata column and a list of mappings (name <- IDs).

• 60 whitelist.subset.ls.Seurat()

  whitelist.subset.ls.Seurat. Subsets cells in a list of Seurat objects based on an externally provided list of cell IDs.

• 61 FindCorrelatedGenes()

  FindCorrelatedGenes. Find correlated genes in a Seurat object

• 62 UpdateGenesSeurat()

  Update Gene Symbols in a Seurat Object. This function updates gene symbols in a Seurat object based on current gene nomenclature guidelines, using HGNChelper(). It checks and updates gene symbols to their latest approved versions,ensuring that gene annotations are current and consistent. The function optionally enforces unique gene symbols and provides statistics on the update process.

• 63 RenameGenesSeurat()

  Rename Gene Symbols in a Seurat Object. This function replaces gene names across various slots within a specified assay of a Seurat object. It is designed to be run prior to any data integration or downstream analysis processes. The function targets the @counts, @data, and @meta.features slots within the specified assay, ensuring consistency in gene nomenclature across the object.

• 64 .check_and_rename()

  Check and Rename Gene Names in Seurat Assay Object. This function renames rows (genes) in a specified slot of a Seurat assay object. It supports slots storing data as either a dense or a sparse matrix (dgCMatrix) or data.frame.

• 65 RemoveGenesSeurat()

  Remove Specific Genes from a Seurat Object. Removes specified genes from the metadata, counts, data, and scale.data slots of a Seurat object. This operation is typically performed prior to data integration to ensure that gene sets are consistent across multiple datasets. The function modifies the Seurat object in place.

• 66 HGNC.EnforceUnique()

  Enforce Unique HGNC Gene Symbols. Ensures that gene symbols are unique after being updated with HGNC symbols. This function applies a suffix to duplicate gene symbols to enforce uniqueness. While using make.unique might not be the ideal solution due to potential mismatches, it significantly reduces the number of mismatching genes in certain scenarios, making it a practical approach for data integration tasks.

• 67 GetUpdateStats()

  Gene Symbol Update Statistics. Generates statistics on the gene symbol updates performed by UpdateGenesSeurat(). This function analyzes the data frame of gene symbols before and after the update process, providing insights into the proportion and total number of genes that were updated.

• 68 PlotUpdateStats()

  PlotUpdateStats. Creates a scatter plot of update statistics.

• 69 Convert10Xfolders()

  Convert10Xfolders. This function takes a parent directory with a number of subfolders, each containing the standard output of 10X Cell Ranger. It (1) loads the filtered data matrices, (2) converts them to Seurat objects, and (3) saves them as .RDS files.

• 70 ConvertDropSeqfolders()

  ConvertDropSeqfolders. This function takes a parent directory with a number of subfolders, each containing the standard output of 10X Cell Ranger. It (1) loads the filtered data matrices, (2) converts them to Seurat objects, and (3) saves them as .RDS files.

• 71 LoadAllSeurats()

  LoadAllSeurats. This function loads all Seurat objects found in a directory. It also works with symbolic links (but not with aliases).

• 72 read10x()

  Load 10X Genomics Data as Seurat Object. Reads 10X Genomics dataset files (gzipped) including matrix, features, and barcodes, to a single expression matrix. This function handles the unzipping of these files, reads the data, and re-compresses the files back to their original gzipped format.

• 73 .saveRDS.compress.in.BG()

  .saveRDS.compress.in.BG. Save and RDS object and compress resulting file in the background using system(gzip). OS X or unix.

• 74 isave.RDS()

  isave.RDS. Save an RDS object, using a faster and efficient compression method that runs in the background.

• 75 xsave()

  Save an R Object Using 'qs' Package for Fast Compressed Saving. This function saves an R object to a file in a quick and efficient format using the 'qs' package. It constructs the file name based on various inputs and stores additional metadata if the object is a Seurat object. The saving path can be adjusted by the presence of 'OutDir' in the global environment or defaults to the working directory.

• 76 xread()

  Read an R Object Using 'qs' Package for Fast Decompression. This function reads an R object from a file saved in a format specific to the 'qs' package, which is designed for quick and efficient compression and decompression of R objects. It also times the read operation, providing feedback on the duration of the operation.

• 77 .getCPUsCBE()

  Get the number of CPUs to use for CBE processing. This function checks for the presence of a global CBE.params list and, if found and contains a cpus entry, returns the number of CPUs specified by cpus minus one. Otherwise, it returns a default number of CPUs.

• 78 isave.image()

  isave.image. Save an image of the current workspace using a faster and efficient compression method that runs in the background.

• 79 qsave.image()

  Save workspace - qsave.image. Faster saving of workspace, and compression outside R, when it can run in the background. Seemingly quite CPU hungry and not very efficient compression. #

• 80 # findBamFilesInSubdirs()

  Find 'Outs' Subdirectories in Specified Subdirectories. This function searches through specified subdirectories within a root directory to find all subdirectories named 'outs' and returns a character vector with their full paths.

• 81 clip10Xcellname()

  Clip Suffixes from 10X Cell Names. Removes suffixes from cell names that are added by 10X technology and Seurat during data processing.

• 82 make10Xcellname()

  Add Suffix to Cell Names (e.g. lane suffix: _1). Appends a specified suffix to cell names to mimic lane suffixes used in 10X datasets.

• 83 plotTheSoup()

  plotTheSoup. Plot stats about the ambient RNA content in a 10X experiment.

• 84 jJaccardIndexVec()

  jJaccardIndexVec. Calculate jaccard similarity for 2 vecotrs. Helper to jPairwiseJaccardIndexList.

• 85 jPairwiseJaccardIndexList()

  jPairwiseJaccardIndexList. Create a pairwise jaccard similarity matrix across all combinations of columns in binary.presence.matrix. Modified from: https://www.displayr.com/how-to-calculate-jaccard-coefficients-in-displayr-using-r/

• 86 jPresenceMatrix()

  jPresenceMatrix. Make a binary presence matrix from a list. Source: https://stackoverflow.com/questions/56155707/r-how-to-create-a-binary-relation-matrix-from-a-list-of-strings #

• 87 jJaccardIndexBinary()

  jJaccardIndexBinary. Calculate Jaccard Index. Modified from: https://www.displayr.com/how-to-calculate-jaccard-coefficients-in-displayr-using-r/ #

• 88 jPairwiseJaccardIndex()

  jPairwiseJaccardIndex. Create a pairwise jaccard similarity matrix across all combinations of columns in binary.presence.matrix. Modified from: https://www.displayr.com/how-to-calculate-jaccard-coefficients-in-displayr-using-r/

• 89 compareVarFeaturesAndRanks()

  Compare variable features and their ranks in two Seurat objects.. This function compares variable features (genes) between two Seurat objects, reporting the number of genes in each, the percentage of common genes, the percentage of unique genes in each object, and the similarity in the ranking of overlapping genes using Spearman's rank correlation coefficient. Optionally, it can also generate a scatterplot of the ranks of common genes using ggpubr's ggscatter. The function returns the common genes and the Spearman's rank correlation coefficient.

• 90 processSeuratObject()

  Process Seurat Objects in Parallel. Applies a series of Seurat processing steps to each Seurat object in a list. The operations include scaling data, running PCA, UMAP, finding neighbors, and finding clusters. This is done in parallel using multiple cores.

• 91 regress_out_and_recalculate_seurat()

  Regress Out and Recalculate Seurat. The function performs a series of calculations and manipulations on a Seurat object, including identifying variable features, scaling data, running PCA, setting up reductions, finding neighbors, and finding clusters. It optionally performs t-SNE and saves the object.

• 92 .getRegressionVariablesForScaleData()

  Check List Elements. Tests if list elements are defined and reports the value or issues a warning.

• 93 .parseKeyParams()

  Parse key parameters from an object and format as a string. This function extracts the number of scaled features, the number of principal components, and formats additional information including regression variables.

• 94 removeScaleData()

  Parse basic obj stats. Parse cell and feature count to a string.

List of Functions in Seurat.Utils.Visualization.R (57)

Updated: 2024/03/22 15:41

• 1 PlotFilters()

  PlotFilters. Plot filtering threshold and distributions, using four panels to highlight the relation between Gene- and UMI-count, ribosomal- and mitochondrial-content.

• 2 scCalcPCAVarExplained()

  Calculate the percent of variation explained by individual PC's. This function calculates the percentage of variation each principal component (PC) accounts for in a Seurat object. It's specifically tailored for Seurat objects and provides a convenient way to understand the variance distribution across PCs. For similar calculations on standard PCA objects, refer to github.com/vertesy/Rocinante PCA.percent.var.explained().

• 3 scPlotPCAvarExplained()

  Plot the percent of variation explained by individual PC's. This function plots the percentage of variation explained by each principal component (PC) in a Seurat object. It allows for a visual assessment of how much variance is captured by each PC, which is crucial for dimensionality reduction analyses. Users can choose between two plotting methods: one using MarkdownReports and the other using ggExpress.

• 4 PercentInTranscriptome()

  Gene Expression as Fraction of Total UMI Counts. This function computes and visualizes gene expression levels as a fraction of total UMI (Unique Molecular Identifier) counts across all genes in a Seurat object. It aims to highlight the relative contribution of the most highly expressed genes to the overall transcriptome.

• 5 geneExpressionLevelPlots()

  Histogram of Gene Expression Levels. This function generates a histogram to visualize the expression level distribution of a specified gene across all cells in a Seurat object. It highlights the position of the gene of interest within the overall distribution.

• 6 PrctCellExpringGene()

  Proportion of Cells Expressing Given Genes. Calculates the proportion of cells expressing one or more specified genes.

• 7 ww.calc_helper()

  Helper to calculate Cell Expression Proportion for Gene. Computes the proportion of cells expressing a specific gene within a Seurat object.

• 8 get.clustercomposition()

  Cluster Composition Analysis. Analyzes and visualizes the composition of clusters in a Seurat object, indicating the contribution of different datasets to each cluster.

• 9 scBarplot.CellFractions()

  Generate Barplot of Cell Fractions. This function generates a bar plot of cell fractions per cluster from a Seurat object. It offers the option to downsample data, equalizing the number of cells in each group to the number in the smallest group. The plot's bars are grouped by one variable and filled by another. The function supports custom color palettes, drawing numerical values on bars, and saving the plot.

• 10 scBarplot.CellsPerCluster()

  Barplot of Fraction of Cells per Cluster. Visualizes the fraction of cells within each cluster through a barplot.

• 11 scBarplot.CellsPerObject()

  Barplot of Cells Per Seurat Object. Visualizes the number of cells in each Seurat object within a list, showing the distribution of cell counts across different datasets or experimental conditions.

• 12 plotClustSizeDistr()

  Cluster Size Distribution Plot (Barplot or Histogram). Generates a bar plot or histogram to visualize the size distribution of clusters within a Seurat object, based on the specified clustering identity.

• 13 scBarplot.FractionAboveThr()

  Barplot the Fraction of Cells Above Threshold per Cluster. Generates a bar plot depicting the percentage of cells within each cluster that exceed a specified threshold, based on a selected metadata column.

• 14 scBarplot.FractionBelowThr()

  Fraction of Cells Below Threshold per Cluster. Generates a bar plot to visualize the percentage of cells within each cluster that fall below a specified threshold, according to a metadata column value.

• 15 scBarplotStackedMetaCateg_List()

  Stacked Barplot of Metadata Categories for List of Seurat Objects. Creates and saves a stacked barplot for a specified metadata category from a list of Seurat objects.

• 16 gg_color_hue()

  Reproduce the ggplot2 default color palette. Generates a vector of colors that emulates the default color palette used by ggplot2. This function is useful for creating color sets for custom plotting functions or for applications outside of ggplot2 where a similar aesthetic is desired.

• 17 # getDiscretePalette()

  Safely generate a discrete color palette (NA).. Safe wrapper around Seurat's DiscretePalette(), which returns NA's if too many categories are requested

• 18 getDiscretePaletteObj()

  Generate a Discrete Color Palette for Seurat Clusters. Generates a discrete color palette for visualizing clusters in a Seurat object, using a specified identity column to determine the number of unique clusters.

• 19 DiscretePaletteSafe()

  Safely generate a Discrete color palette.. Generates a discrete color palette, ensuring no NA values are included, suitable for visualizations where a specific number of distinct, reproducible colors is needed.

• 20 getClusterColors()

  Regenerate Cluster Colors from a Seurat Object. Regenerate and optionally displays the color scheme associated with the clusters in a Seurat object as defined by a specified identity column.

• 21 SeuratColorVector()

  Regenerate Color Scheme for Clusters in Seurat Object as a vector. Extracts and optionally displays the color scheme assigned to cluster identities within a Seurat object, facilitating consistent color usage across visualizations. You can check results in a barplot with MarkdownHelpers::color_check().

• 22 plotAndSaveHeatmaps()

  Plot and Save Heatmaps from Metadata Calculation Results. Generates and saves heatmap visualizations for each metric in the results obtained from metadata calculations, such as mean or median values of specified features across different categories.

• 23 qFeatureScatter()

  Scatter Plot of Two Features in Seurat Object. Generates a scatter plot comparing two features (genes or metrics) from a Seurat object and optionally saves it. The function wraps around Seurat's FeatureScatter for enhanced usability, including optional logarithmic transformations and saving capabilities.

• 24 qSeuViolin()

  Create a Violin Plot for a Seurat Object Feature and save the file.. Generates a violin plot for a specified feature in a Seurat object, allowing for the data to be split by a specified grouping variable. The function supports customization options such as logarithmic scaling, custom titles, and more.

• 25 plotGeneExpHist()

  Histogram of Gene Expression in Seurat Object. Creates and optionally saves a histogram showing expression levels of specified genes within a Seurat object. Provides options for aggregate gene expression, expression threshold filtering, and quantile clipping for count data.

• 26 qUMAP()

  Quick UMAP Visualization of Gene Expression and automatically save the plot. Generates a UMAP visualization for a specific feature from a Seurat object, and automatically saves it. Offers options for custom titles, subtitles, saving, and more. Assumes default options for custom titles, subtitles, saving, and more.

• 27 clUMAP()

  Quick Visualization of Clustering Results with UMAP and automatically save the plot. Generates a UMAP visualization based on clustering results from a Seurat object, and automatically saves it. Offers options for custom titles, subtitles, saving, and more. Assumes default options for custom titles, subtitles, saving, and more.

• 28 umapHiLightSel()

  Highlight Selected Clusters on UMAP. Generates a UMAP plot from a Seurat object with specified clusters highlighted. It saves the resulting UMAP plot directly to the current working directory.

• 29 DimPlot.ClusterNames()

  DimPlot.ClusterNames. Plot UMAP with Cluster names.

• 30 multiFeaturePlot.A4()

  multiFeaturePlot.A4. Save multiple FeaturePlots, as jpeg, on A4 for each gene, which are stored as a list of gene names.

• 31 multiSingleClusterHighlightPlots.A4()

  Generate Cluster Highlight UMAPs compiled into A4 pages. This function generates and saves cluster highlight plots for both single and multiple clusters using UMAP or other dimensionality reduction techniques. It supports saving plots in various formats and allows customization of plot appearance and layout.

• 32 multiSeuratHeatmap.A4()

  multiSeuratHeatmap.A4. Save multiple FeatureHeatmaps from a list of genes on A4 jpeg.

• 33 multi_clUMAP.A4()

  Plot multiple categorical variables in combined UMAPs. Generates and saves multiple UMAP plots for clustering results, adjusting the layout and plot dimensions. Supports the generation of plots in different formats and customization of the visual appearance.

• 34 qClusteringUMAPS()

  Quick Clustering UMAPs on A4 Page. Generates and arranges UMAP plots for up to four specified clustering resolutions from a Seurat object onto an A4 page, facilitating comparative visualization.

• 35 plotQUMAPsInAFolder()

  Plot qUMAPs for Genes in a Folder. This function plots qUMAPs for a specified set of genes, storing the results in a specified folder. If no folder name is provided, it defaults to using the gene set name.

• 36 PlotTopGenesPerCluster()

  Plot Top N Differentially Expressed Genes Per Cluster. Visualizes the top N differentially expressed (DE) genes for each cluster within a specified clustering resolution of a Seurat object, facilitating the exploration of gene expression patterns across clusters.

• 37 qQC.plots.BrainOrg()

  Quickly Plot Key QC Markers in Brain Organoids. Generates and arranges UMAP plots for specified QC features from a Seurat object on an A4 page, facilitating a quick quality control (QC) overview.

• 38 qMarkerCheck.BrainOrg()

  Quickly Plot Key Markers in Brain Organoids. Generates plots for a predefined or custom set of gene markers within brain organoids, aiding in the quick assessment of their expression across different cells or clusters.

• 39 PlotTopGenes()

  Plot Top Genes. This function plots the highest expressed genes on UMAPs, saving the plots in a subfolder. It requires the prior execution of calc.q99.Expression.and.set.all.genes.

• 40 FlipReductionCoordinates()

  Flip Reduction Coordinates. Flips dimensionality reduction coordinates (such as UMAP or tSNE) vertically or horizontally to change the visualization perspective.

• 41 AutoNumber.by.UMAP()

  Relabel Cluster Numbers Along a UMAP (or tSNE) Axis. Automatically renumbers clusters based on their position along a specified dimension in a UMAP (or tSNE or PCA) plot, potentially enhancing interpretability by ordering clusters.

• 42 .adjustLayout()

  Adjust Layout Parameters for multi* plotting fucntions. Adjusts layout dimensions and properties based on the specified layout type. Updates the provided environment with new dimensions and layout configuration.

• 43 save2plots.A4()

  Save Two Plots on One A4 Page. Arranges and saves two UMAP plots (or any plots) side-by-side or one above the other on a single A4 page.

• 44 save4plots.A4()

  Save Four Plots on One A4 Page. Arranges and saves four plots (e.g. UMAPs) onto a single A4 page, allowing for a compact comparison of different visualizations or clustering results.

• 45 qqSaveGridA4()

  qqSaveGridA4. Saves a grid of 2 or 4 ggplot objects onto an A4 page.

• 46 ww.check.if.3D.reduction.exist()

  ww.check.if.3D.reduction.exist. ww.check.if.3D.reduction.exist in backup slot #

• 47 ()

  ww.check.quantile.cutoff.and.clip.outliers. Function to check a specified quantile cutoff and clip outliers from a given expression vector.

• 48 plot3D.umap.gene()

  plot3D.umap.gene. Plot a 3D umap with gene expression. Uses plotly. Based on github.com/Dragonmasterx87.

• 49 plot3D.umap()

  plot3D.umap. Plot a 3D umap based on one of the metadata columns. Uses plotly. Based on github.com/Dragonmasterx87.

• 50 SavePlotlyAsHtml()

  SavePlotlyAsHtml. Save a Plotly 3D scatterplot as an HTML file.

• 51 BackupReduction()

  Backup Dimensionality Reduction Data. Stores a backup of specified dimensionality reduction data (e.g., UMAP, tSNE, PCA) within the Seurat object, from obj@reductions$umap to the @misc$reductions.backup slot.

• 52 SetupReductionsNtoKdimensions()

  Compute and Backup Dimensionality Reductions. Executes specified dimensionality reduction (UMAP, tSNE, or PCA) over a range of dimensions and backs up the results within a Seurat object. This function allows for exploration of data structure at varying levels of granularity and ensures that reduction results are preserved for future reference.

• 53 RecallReduction()

  Recall Dimensionality Reduction from backup slot. Restores dimensionality reduction data (e.g., UMAP, tSNE, PCA) from a backup stored within obj@misc$reductions.backup to the active obj@reductions slot.

• 54 Annotate4Plotly3D()

  Annotate4Plotly3D. Create annotation labels for 3D plots. Source https://plot.ly/r/text-and-annotations/#3d-annotations.

• 55 Plot3D.ListOfGenes()

  Plot3D.ListOfGenes. Plot and save list of 3D UMAP or tSNE plots using plotly.

• 56 Plot3D.ListOfCategories()

  Plot3D.ListOfCategories. This function plots and saves a list of 3D UMAP or tSNE plots using plotly.

• 57 # panelCorPearson()

  Display Correlation Values in Pairs Plot. This function displays the correlation coefficient and significance level within a scatterplot generated by the pairs() function. The default correlation method is Pearson, but Kendall or Spearman methods can also be selected.

List of Functions in Seurat.Utils.Metadata.R (27)

Updated: 2024/03/22 15:41

• 1 getMetaColnames()

  Get Metadata Column Names Matching Pattern. Retrieves column names from an object's metadata that match a specified pattern.

• 2 metaColnameExists()

  Check if a Column Exists in the Metadata of an S4 Object. This function checks whether a given column exists in the meta.data of a Seurat object.

• 3 getMetadataColumn()

  getMetadataColumn. Retrieves a specified metadata column from a Seurat object and returns it as a named vector.

• 4 get_levels_seu()

  Get Unique Levels of a Seurat Object Ident Slot. This function extracts the unique levels present in the 'ident' slot of a Seurat object. The function throws an error if the number of levels exceeds 'max_levels'. The function optionally prints the R code to recreate the 'Levels' vector using 'dput'.

• 5 calculateAverageMetaData()

  Calculate Average Metadata for Seurat Object. Computes specified metrics (e.g., median, mean) for given metadata features across each category defined by an identity column in a Seurat object's metadata. This function allows for flexible metric calculation on specified features, providing insights into the data distribution.

• 6 getMedianMetric.lsObj()

  getMedianMetric.lsObj. Get the median values of different columns in meta.data, can iterate over a list of Seurat objects.

• 7 getCellIDs.from.meta()

  getCellIDs.from.meta. Retrieves cell IDs from a specified metadata column of a Seurat object, where the cell ID matches a provided list of values. The matching operation uses the %in% operator.

• 8 addMetaDataSafe()

  Add Metadata to a Seurat object, safely with Checks. Wrapper function for AddMetaData that includes additional checks and assertions.

• 9 seu.add.meta.from.vector()

  seu.add.meta.from.vector. Adds a new metadata column to a Seurat object.

• 10 create.metadata.vector()

  Create a Metadata Vector. This function creates a metadata vector from an input vector and a Seurat object. The resulting vector contains values from 'vec' for the intersecting cell names between 'vec' and 'obj'. It also checks if the intersection between the cell names in 'vec' and 'obj' is more than a minimum intersection size.

• 11 addMetaFraction()

  addMetaFraction. Add a new metadata column to a Seurat object, representing the fraction of a gene set in the transcriptome (expressed as a percentage).

• 12 add.meta.tags()

  add.meta.tags. Add metadata tags to a Seurat object dataset.

• 13 seu.add.meta.from.table()

  seu.add.meta.from.table. Add multiple new metadata columns to a Seurat object from a table. #

• 14 seu.map.and.add.new.ident.to.meta()

  seu.map.and.add.new.ident.to.meta. Adds a new metadata column to a Seurat object based on an identity mapping table.

• 15 fix.orig.ident()

  fix.orig.ident. Remove the string "filtered_feature_bc_matrix." from "orig.ident". Helper function.

• 16 seu.RemoveMetadata()

  seu.RemoveMetadata. Remove specified metadata columns from a Seurat object.

• 17 saveLsSeuratMetadata()

  Save Metadata from a List of Seurat Objects. This function takes a list of Seurat objects, extracts their metadata, and saves it to a file with a specified suffix.

• 18 # transferMetadataV1()

  Transfer Multiple Metadata Columns Between Two Seurat Objects. Transfers specified metadata columns from one Seurat object to another, with options for verbose output and overwriting existing columns. Checks for cell overlap and reports percentages of matching and unique cells.

• 19 sampleNpc()

  Sample N % of a dataframe (obj@metadata), and return rownames (cell IDs).. This function samples a specified percentage of a dataframe (specifically a subset of the metadata of a Seurat object) and returns the corresponding cell IDs.

• 20 writeCombinedMetadataToTsvFromLsObj()

  Combine Metadata from a list of Seurat objects and Write to TSV. Formerly writeMetadataToTsv. writeCombinedMetadataToTsvFromLsObj takes a list of ls.Obj, extracts their @meta.data slots, removes specified columns, checks for column consistency, creates a barplot showing the number of rows per object, and finally merges these into one large data frame.

• 21 plotMetadataCorHeatmap()

  Plot Metadata Correlation Heatmap. This function plots a heatmap of metadata correlation values. It accepts a Seurat object and a set of metadata columns to correlate. The correlations are calculated using either Pearson or Spearman methods, and the resulting heatmap can include the principal component (PCA) values and be saved with a specific suffix.

• 22 heatmap_calc_clust_median()

  Calculate and plot heatmap of cluster medians. This function calculates the median of specified variables in a dataframe, grouped by a column ('ident'). The function also provides an option to scale the medians, subset the ident levels, and either return a matrix of median values or plot a heatmap.

• 23 plotMetadataCategPie()

  plotMetadataMedianFractionBarplot. Generates a barplot of metadata median values.

• 24 renameAzimuthColumns()

  Rename Azimuth Columns in Seurat Object. Dynamically renames specified metadata columns in a Seurat object, particularly those prefixed with "predicted." and the "mapping.score" column, by applying a new prefix that combines a user-defined prefix and a reference name.

• 25 renameSmallCategories()

  Rename Small Categories in Seurat Object Metadata. This function renames categories within a specified identity column of a Seurat object's metadata that have fewer cells than a specified minimum threshold. Categories below this threshold are renamed to a common name, typically "unclear", to clean up small, potentially noisy categories.

• 26 .metaColnames()

  Transfer labels from a reference Seurat object to a query Seurat object. Function to transfer labels from a reference Seurat object to a query Seurat object using anchoring and transfer data methods from the Seurat package. It then visualizes the reference and the combined objects using Uniform Manifold Approximation and Projection (UMAP).

• 27 matchBestIdentity()

  Match and Translate Best Identity. This function matches the best identity from ident_to_rename to reference_ident in an object, in other words, it replaces original categories with the most frequent ones from the reference, hence helps to filter out less important categories.

List of Functions in Seurat.utils.less.used.R (5)

Updated: 2024/03/22 15:41

• 1 plot.UMAP.tSNE.sidebyside()

  plot.UMAP.tSNE.sidebyside. Plot a UMAP and tSNE side by side.

• 2 umapNamedClusters()

  Plot and Save UMAP without legend. Generates a UMAP plot colored by a specified metadata column and saves the plot to a file.

• 3 AutoNumber.by.PrinCurve()

  AutoNumber.by.PrinCurve. Relabel cluster numbers along the principal curve of 2 UMAP (or tSNE) dimensions. #

• 4 load10Xv3()

  Load 10X Genomics Version 3 Data. Loads 10X Genomics data from a specified directory containing output folders for raw and filtered data. This function is designed to handle data from 10X Genomics Chromium Single Cell technologies (version 3).

• 5 Convert10Xfolders.old()

  Convert10Xfolders.old. This function takes a parent directory with a number of subfolders, each containing the standard output of 10X Cell Ranger. It (1) loads the filtered data matrices, (2) converts them to Seurat objects, and (3) saves them as .RDS files.

Usage

You can use most functions at relevant steps of a standard Seurat analysis.

We are preparing a vignette.

---

Get Seurat.utils. Vertesy, 2024. If you use these functions, please star the repo, or cite via DOI. Thanks!

seurat.utils's Issues

qUMAP should have auto-suffixing by splitby

• qUMAP should have auto-suffixing by splitby
• rename splitby to split.by if does not break

GetNamedClusteringRuns should return non-named ones if named ones are not found

Will break much less use case

How to make seurat object smaller?

Ideas

• Remove stuff (not preferred), see DietSeurat.
• Round up scale data and normdata slots.

PlotFilters should handle solo objects not only lists of objects.

Now

 ls.Seurat <- list('a' = combined.obj)
  PlotFilters(ls.obj = ls.Seurat)

Future

  PlotFilters(ls.obj = combined.obj)

PlotGoTermScores shows equal plots for very different gene sets

Observation

PlotGoTermScores() shows equal plots for very different gene sets. See equalp plots and very diff. gene set sizes:

GO 0001705 ectoderm formation | 4 vs 1 gene

GO.0007498

Repeatedly true. This cannot be.

---

## PlotGoTermScores
function(obj = combined.obj # Automate retrieving, processing and plotting GO term based gene scores.
                             , openBrowser = F
                             , GO = "GO:0061621", desc = "canonical.glycolysis") {
  GO.wDot<- make.names(GO)

  obj <- GetGOTerms(obj = obj, GO = GO, web.open = openBrowser);
  iprint(desc, obj@misc$GO[[ GO.wDot ]])
  obj <- AddGOScore(obj = obj, GO = GO);
  FeaturePlotSave(obj = obj, GO = paste0("Score.", GO.wDot), name_desc = desc)
  return(obj)
}

Correct:

1. GetGOTerms()
2. Part of AddGOScore() -> stored in the list

But is: AddModuleScore correct?

AutoLabel.KnownMarkers returns only 1 cluster, when AutoLabelTop.logFC(res = res) : Not all clusters returned DE-genes!

Describe the bug
AutoLabelTop.logFC(res = res) : Not all clusters returned DE-genes!

Expected behavior
Make up names for missing

Additional context

> combined.obj <- StoreAllMarkers(df_markers = df.markers, res = res)
[1] "DF markers are stored under: obj@misc$df.markers$ res.0.3"
>       combined.obj <- AutoLabelTop.logFC(res = res);
Adding missing grouping variables: `cluster`
[1] "missing: cl.names.KnownMarkers.0.3"
[1] "top.markers-umap/"
[1] "All files will be saved under 'NewOutDir':  ~/Dropbox/Abel.IMBA/AnalysisD/SEO/scTE.retrotransposon.analysis.test.CON.sc/top.markers-umap/"
[1] "ParentDir was defined as: ~/Dropbox/Abel.IMBA/AnalysisD/SEO/scTE.retrotransposon.analysis.test.CON.sc/Differential.Gene.expression/res.0.3/"
[1] "ParentDir will be: ~/Dropbox/Abel.IMBA/AnalysisD/SEO/scTE.retrotransposon.analysis.test.CON.sc/"
[1] "Call *create_set_Original_OutDir()* when chaning back to the main dir."
[1] "1 or 6.67 % genes not found in the data, e.g: cl.names.KnownMarkers.0.3"
[1] "layout active, nr.Col ignored."
[1] "1 LTR7 AC009133.6 AP000812.5"
[1] "2 COL3A1 DLX6-AS1 VIM TXNDC5"
[1] "3 STMN2 L4-C-Mam CCNI"
[1] "4 MT-TF PTN AC020613.1"
[1] "All files will be saved under 'OutDir':  ~/Dropbox/Abel.IMBA/AnalysisD/SEO/scTE.retrotransposon.analysis.test.CON.sc/"
5.736 sec elapsed
Warning message:
In AutoLabelTop.logFC(res = res) : Not all clusters returned DE-genes!

quantile cutoff in plot3D.umap.gene works differently than in DimPlot

• It seems that quantile cutoff in DimPlot() works nicer than in plot3D.umap.gene().
• I assume it's because they cutoff differently: probably only on the non-zero values
• Not verified, but once it is , we should implement it

PlotFilters gives back a meaningless filename suffix

What is a.. in Filtering.thresholds.a..png?

Last gene is lost in qUMAP (repeats first gene in plot title)

Genes.panel.C <- c('PDK1', 'ENO1', 'LDHA', 'VEGFA', 'DDIT3', 'XBP1', 'DDIT4', 'P4HB')
qUMAP(feature = Genes.panel.C, nr.cols = 4, w = 10)

generates

It looks like that only title is messed up

clUMAP freezes R if you (mistakenly) call it on a continuous variable

• It should give a warning if too many "categories" detected
• It should not execute if "categories" > X

Since Roxygen3, some functions are exported as S3 generics, and cannot be found

See: vertesy/CodeAndRoll2#9

# Generated by roxygen2: do not edit by hand

S3method(barplot,cells.per.cluster)
S3method(plot,Gene.Cor.Heatmap)
S3method(plot,Metadata.Cor.Heatmap)
S3method(plot,Metadata.median.fraction.barplot)
S3method(plot,UMAP.tSNE.sidebyside)
S3method(plot,clust.size.distr)
S3method(plot,expression.rank.q90)
S3method(points_on_curve,SlingshotDataSet)
S3method(points_on_curve,principal_curve)

• Add an explicit #' @export sort.decreasing to solve the problem

recall.all.genes() spams the console

Problem after all.genes defined as: 1 0.999

recall.all.genes()
   MALAT1    TMSB4X    TMSB10    TUBA1A    MT-CO1     RPS19 
1.0000000 0.9999622 0.9999244 0.9998865 0.9998487 0.9998109 
[1] "all.genes defined as: 1 0.999962175656252 0.999924351312505 0.999886526968757 0.999848702625009 0.999810878281262 0.999773053937514 0.999735229593767 0.999697405250019 0.999659580906271 0.999621756562524 0.999583932218776 0.999508283531281 0.999508283531281 0.999470459187533 0.999432634843785 0.999394810500038 0.999319161812543 0.999319161812543 0.999281337468795 0.999167864437552 0.999167864437552 0.999167864437552 0.999130040093804 0.999054391406309 0.999054391406309 0.999016567062561 0.998978742718814 0.998940918375066 0.998903094031319 0.998865269687571 0.998827445343823 0.998789621000076 0.998751796656328 0.99871397231258 0.998676147968833 0.998638323625085 0.998600499281337 0.99856267493759 0.998487026250095 0.998487026250095 0.998449201906347 0.998411377562599 0.998297904531356 0.998297904531356 0.998297904531356 0.998260080187609 0.998222255843861 0.998184431500113 0.998146607156366 0.998108782812618 0.998033134125123 0.998033134125123 0.997957485437628 0.997957485437628 0.997881836750132 0.997881836750132 0.997844012406385 0.997806188062637 0.997768363718889 0.997730539375142 0.997579242000151 0.997579242000151 0.997579242000151 0.997579242000151 0.997541417656404 0.997503593312656 0.997427944625161 0.997427944625161 0.997314471593918 0.997314471593918 0.997314471593918 0.99727664725017 0.997238822906423 0.99712534987518 0.99712534987518 0.99712534987518 0.997087525531432 0.997049701187684 0.996936228156441 0.996936228156441 0.996936228156441 0.99

3D plotting should work without pre-setting anything

plot3D.umap and plot3D.umap.gene() both require RecallReduction() before to get the 3D umap cooridnated.

It should be done internally.

saveRDS.compress.in.BG does not execute gzip command

saveRDS.compress.in.BG <- function(obj, compr = FALSE, fname) {
  try(tictoc::tic(), silent = T)
  saveRDS(object = obj, compress = compr, file = fname)
  try(tictoc::toc(), silent = T)
  print(paste("Saved, being compressed", fname))
  system(paste("gzip", fname),  wait = FALSE) # execute in the background
  try(say(), silent = T)
}

New Recall.Seurat.Analysis function

Write a Recall.Seurat.Analysis function, that loads all the things you need to continue the analysis

Recall.Seurat.Analysis <- function(path2obj = "/Users/abel.vertesy/Dropbox/Group Folder Knoblich/Users/Abel.dshared/abel.camilla/2020.10.01/POL.sc.preAnalysis.premRNA./combined.obj__2020.10.01_22.26.Rds.gz"
                            , required.packages = list('tibble', 'ggplot2', 'readr', 'dplyr', 'cowplot', 'stringr', 'Seurat', 'MarkdownReportsDev', 'tictoc', 'schex')
                            ) {

  iprint("Loading packages... ------------------------------------------------")
  lapply(required.packages, require, character.only = TRUE)

  # Try sourcing standalone scripts ---------------------
  iprint("Loading custom scripts... ------------------------------------------------")
  # try(source ('~/GitHub/Packages/CodeAndRoll/CodeAndRoll.R'))
  # try(source("~/GitHub/Packages/Seurat.utils/Seurat.Utils.Load.R"));
  # try(source("~/GitHub/Packages/Seurat.multicore/Seurat3.Multicore.Load.R"));

  print("Loading Seurat object... ------------------------------------------------")
  combined.obj <- read_rds(path2obj)
  
  print("Loading Seurat object... ------------------------------------------------")
  assign(x = combined.obj, value = "combined.obj", envir = as.environment(x = 1))
  # qUMAP("MAP2")
  recall.all.genes()
  recall.parameters()
  

}
# Recall.Seurat.Analysis()

calc.cluster.averages does should have custom suffixing and adding the split_by (res) to the file name

Maybe even the qplot* functions should have this?

ToDo

ToDo list

PlotGoTermScores does not find enough genes for large GO terms

PlotGoTermScores() does not find enough genes for large GO terms:

For example response to stress GO:0006950

PlotGoTermScores(GO = "GO:0006950", desc= "response to stress", obj = combined.obj)
[1] "0 Gene symbols downloaded:"
[1] "0 genes (of 0 ) are MISSING from the Seurat object:"
[1] "Genes in GO:0006950 are saved under obj@misc$GO$ GO.0006950"
[1] "response to stress"
Error in AddModuleScore(object = obj, features = genes.GO, name = ScoreName) : 
  The following feature lists do not have enough features present in the object:  exiting...
In addition: Warning message:
In AddModuleScore(object = obj, features = genes.GO, name = ScoreName) :
 
 Error in AddModuleScore(object = obj, features = genes.GO, name = ScoreName) : 
  The following feature lists do not have enough features present in the object:  exiting... 

Tracing

1. Problem goes back to GetGOTerms(), and then to
2. getBM() as seen from "0 Gene symbols downloaded:"

> getBM(attributes=c('hgnc_symbol'), #  'ensembl_transcript_id', 'go_id'
+       filters = "go",  uniqueRows = TRUE,
+       values = 'GO:0006950', mart = ensembl)[,1]
logical(0)

The same works for other

getBM(attributes=c('hgnc_symbol'), #  'ensembl_transcript_id', 'go_id'
      filters = "go",  uniqueRows = TRUE,
      values = 'GO:0014002', mart = ensembl)[,1]

Unclear why GO:0006950 returns nothing ensembl mart.

Define multiFeaturePlot.Landscape.A4 for 4 landscape plots

Or maybe define 2 defaults within the same function

Examples of (new) functions' output.

Here I show how example outputs of functions.

PlotFilters should give back the cells surviving / give back the downsampled object directly

Error: Failed to install 'unknown package' from GitHub: HTTP error 403

devtools::install_github(repo = "vertesy/Seurat.utils") 
Error: Failed to install 'unknown package' from GitHub: HTTP error 403. 
API rate limit exceeded for 193.171.188.3. (But here's the good news: 
Authenticated requests get a higher rate limit. Check out the documentation for more details.) 
Rate limit remaining: 0/60 Rate limit reset at: 2021-11-08 12:51:40 UTC To increase your GitHub API rate limit 
- Use `usethis::create_github_token()` to create a Personal Access Token. 
- Use `usethis::edit_r_environ()` and add the token as `GITHUB_PAT`.

recall.parameters() did not work

recall.parameters() did not work when loading AUT using ~GitHub/Projects/AUT/standalone/Glycolysis.genes.R

> recall.parameters()
Error in print(p) : object 'p' not found

Error in clUMAP, when label = F

>   (clUMAP("integrated_snn_res.0.3.Manual.short", label = F))
Rasterizing points since number of points exceeds 100,000.
To disable this behavior set `raster=FALSE`
Error: Can't add `if (!legend) NoLegend() else NULL + if (!axes) NoAxes() else NULL` to a ggplot object.
Run `rlang::last_error()` to see where the error occurred.
>   (clUMAP("integrated_snn_res.0.3.Manual.short"))
Rasterizing points since number of points exceeds 100,000.
To disable this behavior set `raster=FALSE`

Add error bars for calc.cluster.averages

• SEM by default
• Median used, by defaults so Standard Error of the Median should be used by def

PlotFilters() does not fully recap Filtering.plots.R script

The PlotFilters() does not fully recap Filtering.plots.R script.

Seems to be an older/ incomplete version.

PlotFilters should not fail on a list without names OR complain clearly

ls.Seurat <- list('a' = combined.obj)
  PlotFilters(ls.obj = ls.Seurat)


Error in if (suffices == l(ls.obj)) print("ls.obj elements have no names (required).") : 
  argument is of length zero

Read10X() Fails to read in SoupX decont matrices in Convert10Xfolders()

Read10X() Fails to read in SoupX decont matrices in Convert10Xfolders()

count_matrix <- Read10X(pathIN)
Error in dimnamesGets(x, value) : 
  invalid dimnames given for “dgTMatrix” object

Seurat.utils depends on itself - error at installation

ERROR: dependency ‘Seurat.utils’ is not available for package ‘Seurat.utils’ * removing ‘/Library/Frameworks/R.framework/Versions/4.1/Resources/library/Seurat.utils’ 
Warning messages: 
1: package ‘Seurat.utils’ is not available for this version of R 
A version of this package for your version of R might be available elsewhere, see the ideas at https://cran.r-project.org/doc/manuals/r-patched/R-admin.html#Installing-packages 
2: In i.p(...) : installation of package ‘/var/folders/ds/ppssj_7d2wl3qfcf078g6wj00000gr/T//RtmpDTwSht/filebcd073a735a0/Seurat.utils_0.9.8.tar.gz’ had non-zero exit status

SoupX decontamination is not to great

Problems:

1. tSNE seem really bad - probably it does it on the wrong genes.
2. cl specific Markers seem bad:

In 2 biol replicate datasets it finds totally different markers

None are neural markers.

It just seem to find genes for smaller outlier clusters. No wonder, if it uses tf-idf.

It would only work well if we supply it the genes from Seurat DE

Or improve its own DE function

Write into a functiion: df.markers.average.fold.enrichment.R ( SEO/standalone)

SEO/standalone/

https://github.com/vertesy/SEO/issues/118

AutoLabel.KnownMarkers and Add.DE.combined.score could auto detect the name of logFC column

avg_logFC for ln() in x<4 or avg_log2FC in v4

Error in PlotFilters() - Error: 'percentage_formatter' is not an exported object from 'namespace:MarkdownReports'

It seems that a subfunction calls it - maybe even in another package

 PlotFilters(ls.obj = ls.Seurat)
[1] "We filtered for high quality cells based on the number of genes detected [ 500 ; 10000 ] and the fraction of mitochondrial [ 0 % ; 20 % ] and ribosomal [ 0 % ; 30 % ] reads."
[1] "All files will be saved under 'OutDir':  ~/Dropbox/Abel.IMBA/AnalysisD/SEO/Ribo.high.cells/Reanalyze.minFeature.500/Filtering.plots.mito.0.0.2.ribo.0.0.3.nFeature.500.10000./"
[1] "OutDir defined as: ~/Dropbox/Abel.IMBA/AnalysisD/SEO/Ribo.high.cells/Reanalyze.minFeature.500/Filtering.plots.mito.0.0.2.ribo.0.0.3.nFeature.500.10000./"
[1] "a"
 Error: 'percentage_formatter' is not an exported object from 'namespace:MarkdownReports' 

Function parameters are often not documented.

Document most before first non-pre-release.

Error when installing Seurat.utils

Hey Abel,

I loaded all packages w/o problems, except for Seurat.utils

Gave the following error message:

Error: object ‘Stringendo::percentage_formatter’ is not exported by 'namespace:Stringendo'

plot3D.umap.gene cannot plot continous values from meta.data. It is implemented with a bug.

isave should save into the Dropbox folder by default

• isave should save into the Dropbox folder by default,
• complain if not found
• customizable

combined.score in DE does not work well: more weight to log2FC

Problem

In large datasets, p-values plummet and thus have very large influence, over log fold change.

Observations

It still makes sense to use both p value and logFC. In the 3 genes below, there is low difference in FC, but the p-value correctly indicates

Solution

• log of log (p)?
• log (p) / 10?
• log (p) / avg(p)?

getGOTerms without Biomart

Annotate all genes with GO via clusterProfiler:

#Annotate all genes for offline GO visualization (GeneSetScores); own GetGOTerms function
library(clusterProfiler)
enrich.go <- enrichGO(rownames(GetAssayData(combined.obj)), 'org.Hs.eg.db', keyType = "SYMBOL", ont = "ALL", pvalueCutoff=10, qvalueCutoff=10, minGSSize=1, maxGSSize=1000000)
combined.obj@misc$enrichGO[['RNA']] <- enrich.go

If enrichGO object present in "misc" use that otherwise drop back to biomart:

#GetGOTerm used on enrichGO annotation; Any other GO, KEGG, DOSE, ... possible
#enrich.go <- enrichGO(rownames(GetAssayData(combined.obj)), 'org.Hs.eg.db', keyType = "SYMBOL", ont = "ALL", pvalueCutoff=10, qvalueCutoff=10, minGSSize=1, maxGSSize=1000000)
#combined.obj@misc$enrichGO[['RNA']] <- enrich.go
GetGOTerms <- function(obj = combined.obj, GO = 'GO:0034976', web.open = F, genes.shown = 10) { # Get GO terms from enrichGO
print('GetGOTerms()')

if (is.null(obj@misc$enrichGO[['RNA']])) {
while (length(genes) == 0) {
genes <- getBM(attributes = c('hgnc_symbol'), # 'ensembl_transcript_id', 'go_id'
filters = "go_parent_term", uniqueRows = TRUE,
values = GO, mart = ensembl)[,1]
iprint(length(genes), "Gene symbols downloaded:", head(genes, n = genes.shown))
}
} else {
genes <- unlist(geneInCategory(obj@misc$enrichGO[['RNA']])[GO])
iprint(length(genes), "Gene symbols from enrichGO[['RNA']]:", head(genes, n = genes.shown))
}

(GO.wDot <- make.names(GO))

genes <- IntersectWithExpressed(obj = obj, genes = genes)

if (is.null(obj@misc$GO)) obj@misc$GO <- list()
obj@misc$GO[[ GO.wDot ]] <- genes
iprint("Genes in", GO, "are saved under obj@misc$GO$", GO.wDot)
if (web.open) system(paste0("open https://www.ebi.ac.uk/QuickGO/search/", GO))
return(obj)
}

default quantile cutoff in plot3D.umap.gene sometimes cuts away all positive cells. Should be automatically adjusted.

Master recall function

There are quite some variable that are used and stored in the global env.
Upon re-loading a Seurat object, these are missing.

1. I am working on that all are stored in @misc and that all can be individually recalled
2. However a master-recall function could make it easier

Elements to recall

1. p (parameters)
2. gl (gene lists)
3. all.genes
4. meta.tags

Thresholding in scBarplotFractionAboveThr is wrong: count above 0 gives 100% even at

scBarplotFractionAboveThr(columns.used = c('cl.names.top.gene.res.0.3', 'AAV-mCamk2'), thrX = 1)
0%

scBarplotFractionAboveThr(columns.used = c('cl.names.top.gene.res.0.3', 'AAV-mCamk2'), thrX = 0.1)
100%

Temp fix
0.001

qUMAP and others should convert gene names to upper case for convenience

• Which functions should have this?
• What helper would do this inside them?
• Is check.genes equipped for this?
• do qUMAP and others do call check.genes?

getBM() inside GetGOTerms, which is inside PlotGoTermScores() returns no gene.

getBM() inside GetGOTerms, which is inside PlotGoTermScores() returns no gene.

I already had such problem. BioMart seem to change too much. I need to transition to an offline annotation database.

SavePlotlyAsHtml should allow custom suffixing to easily define unique file names

Background color in multiFeaturePlot.A4 turned to black.

Mindblowing: suddenly multiFeaturePlot.A4 BG is black. Cant change it back.

Featureplot is fine.

PlotFilter panels B, C titles should report both upper and loower thr, not just 'Below X'

plot3D.umap.gene should automatically recall the RNA assay, while giving the user a choice

Cluster Auto-naming by highest genes should exclude AL*, AC*, LINC* RNA's

They are not very telling ...

AutoLabelTop.logFC(res = res)

Installation: complains for SoupX dependency. It should not.

Dependencies have multiple problems

• Seurat.utils on SoupX required at install.
• Readwriter on codeandroll required at install.
• ggExpress on Markdownreports required at install.