@@ -1,16 +1,16 @@

 language: r

-cache: 

+cache:

   - packages

   - pip

 r:

-  - oldrel

+  - release

 sudo: false

 warnings_are_errors: true

 bioc_required: true

 # Current version of xgboost fails install

 # Need to force install older version using 'devtools::install_version'

-#    which is done in 'before_install' (see below) 

+#    which is done in 'before_install' (see below)

 # However 'remotes::dev_package_deps' will try to auto upgrade packages

 #    Therefore, a new install command is used with 'upgrade="never"'

 # These parts can be deleted once the xgboost error is fixed

@@ -21,7 +21,7 @@ before_install:

   - Rscript -e 'install.packages("devtools")'

   - Rscript -e 'devtools::install_version("xgboost", version = "0.90.0.2", repos = "http://cran.us.r-project.org")'

   - pip install --upgrade pip

-  - pip install scipy scrublet scanpy scanorama scgen bbknn

+  - pip install scipy scrublet scanpy scanorama bbknn

 install:

 - R -e 'devtools::install_deps(dep = TRUE, upgrade="never")'

@@ -31,11 +31,9 @@ addons:

     packages:

       - libv8-dev

-r_build_args: "--no-build-vignettes"

-r_check_args: "--no-vignettes --no-build-vignettes --as-cran"

+r_check_args: "--as-cran"

 r_github_packages:

   - jimhester/lintr

-  - compbio/celda

 notifications:

   slack:

     secure: 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

@@ -1,6 +1,6 @@

 Package: singleCellTK

 Type: Package

-Title: Interactive Analysis of Single Cell RNA-Seq Data

+Title: Comprehensive and Interactive Analysis of Single Cell RNA-Seq Data

 Version: 1.7.6

 Author: David Jenkins

 Authors@R: c(person(given="David", family="Jenkins", email="dfj@bu.edu", role=c("aut","cre"),

@@ -15,17 +15,14 @@ Authors@R: c(person(given="David", family="Jenkins", email="dfj@bu.edu", role=c(

                     comment = c(ORCID = "0000-0002-6247-6595")))

 Maintainer: David Jenkins <dfj@bu.edu>

 Depends:

-    R (>= 3.6.0),

+    R (>= 4.0),

     SummarizedExperiment,

     SingleCellExperiment,

     DelayedArray,

     Biobase

-Description: Run common single cell analysis directly through your browser

-    including differential expression, downsampling analysis, and clustering.

+Description: Run common single cell analysis in the R console or directly through your browser. Includes many functions for import, quality control, normalization, batch correction, clustering, differential expression, and visualization..

 License: MIT + file LICENSE

 Encoding: UTF-8

-cache:

-  - packages: true

 biocViews: SingleCell, GeneExpression, DifferentialExpression, Alignment,

     Clustering, ImmunoOncology

 LazyData: TRUE

@@ -35,6 +32,7 @@ Imports:

     BiocGenerics,

     BiocParallel,

     colourpicker,

+    colorspace,

     cowplot,

     cluster,

     ComplexHeatmap,

@@ -46,24 +44,25 @@ Imports:

     fields,

     ggplot2,

     ggplotify,

-    ggtree,

     ggrepel,

+    ggtree,

     gridExtra,

     GSVA (>= 1.26.0),

     GSVAdata,

     harmony,

     igraph,

+    KernSmooth,

     liger,

     limma,

     MAST,

     Matrix,

-    mnormt,

     matrixStats,

     methods,

     msigdbr,

     multtest,

     plotly,

     RColorBrewer,

+    ROCR,

     Rtsne,

     S4Vectors,

     scater,

@@ -86,7 +85,6 @@ Imports:

     celda,

     shinycssloaders,

     shinythemes,

-    umap,

     uwot,

     DropletUtils,

     scds (>= 1.2.0),

@@ -94,10 +92,8 @@ Imports:

     tools,

     withr,

     GSEABase,

-    DoubletFinder,

     R.utils,

     zinbwave,

-    V8,

     scRNAseq (>= 2.0.2),

     TENxPBMCData,

     yaml,

@@ -117,12 +113,9 @@ Suggests:

     org.Mm.eg.db,

     stringr

 Remotes: 

-    joshua-d-campbell/DoubletFinder,

     joshua-d-campbell/harmony,

-    rz2333/mnormt@1.5-6,

     joshua-d-campbell/liger,

     joshua-d-campbell/shiny-directory-input,

-    campbio/celda@RELEASE_3_11

 VignetteBuilder: knitr

 URL: https://compbiomed.github.io/sctk_docs/

 BugReports: https://github.com/compbiomed/singleCellTK/issues

@@ -1,5 +1,6 @@

 # Generated by roxygen2: do not edit by hand

+export(MAST)

 export(alignSingleCellData)

 export(calcEffectSizes)

 export(combineSCE)

@@ -8,6 +9,7 @@ export(constructSCE)

 export(convertGeneIDs)

 export(convertSCEToSeurat)

 export(convertSeuratToSCE)

+export(discreteColorPalette)

 export(distinctColors)

 export(downSampleCells)

 export(downSampleDepth)

@@ -16,7 +18,6 @@ export(exportSCE)

 export(exportSCEtoAnnData)

 export(exportSCEtoFlatFile)

 export(featureIndex)

-export(filterSCData)

 export(findMarkerDiffExp)

 export(generateMeta)

 export(generateSimulatedData)

@@ -43,6 +44,7 @@ export(importGeneSetsFromCollection)

 export(importGeneSetsFromGMT)

 export(importGeneSetsFromList)

 export(importGeneSetsFromMSigDB)

+export(importMultipleSources)

 export(importOptimus)

 export(importSEQC)

 export(importSTARsolo)

@@ -86,6 +88,7 @@ export(plotUMAP)

 export(qcInputProcess)

 export(readSingleCellMatrix)

 export(reportCellQC)

+export(reportDiffExp)

 export(reportDropletQC)

 export(reportQCTool)

 export(retrieveSCEIndex)

@@ -114,7 +117,6 @@ export(runMAST)

 export(runMNNCorrect)

 export(runPerCellQC)

 export(runSCANORAMA)

-export(runSCGEN)

 export(runSCMerge)

 export(runScranSNN)

 export(runScrublet)

@@ -160,8 +162,6 @@ import(DelayedArray)

 import(DropletUtils)

 import(GSVAdata)

 import(SingleCellExperiment)

-importFrom(DelayedArray,colSums)

-importFrom(Matrix,colSums)

 importFrom(SingleCellExperiment,"counts<-")

 importFrom(SingleCellExperiment,"reducedDim<-")

 importFrom(SingleCellExperiment,counts)

new file mode 100644

@@ -0,0 +1,136 @@

+#' MAST

+#'

+#' Run and visualize MAST analysis on a SCtkExperiment object.

+#'

+#' @param inSCE Input SCtkExperiment object. Required

+#' @param useAssay The assay to use for the MAST calculations. The default is

+#' "logcounts"

+#' @param condition select variable (from the colData) that is used for the

+#' model.

+#' @param interest.level If the condition of interest has more than two factors,

+#' indicate which level should be used to compare to all other samples.

+#' @param freqExpressed Filter genes that are expressed in at least this

+#' fraction of cells. The default is expression in 0.1 of samples.

+#' @param fcThreshold Minimum fold change for differentially expressed gene.

+#' @param p.value p values for selecting the hurdle result, default is 0.05

+#' @param useThresh Use adaptive thresholding to filter genes. The default is

+#' FALSE.

+#'

+#' @describeIn MAST Run MAST analysis.

+#'

+#' @return MAST(): A data.frame of differentially expressed genes with p-values.

+#' @export

+MAST <- function(inSCE, condition = NULL, interest.level = NULL,

+                 freqExpressed = 0.1, fcThreshold=log2(1.5), p.value = 0.05,

+                 useThresh=FALSE, useAssay = "logcounts"){

+

+  if (is.null(condition)){

+    stop("specify the condition of interest")

+  }

+

+  if (length(unique(SingleCellExperiment::colData(inSCE)[, condition])) == 1) {

+    stop("only one level is in the condition")

+  }

+

+  if (is.null(interest.level) & length(

+    unique(SingleCellExperiment::colData(inSCE)[, condition])) > 2){

+    stop("You must specify a level of interest when more than 2 levels are in",

+         " the condition")

+  }

+

+  # Create MAST SingleCellAssay

+  pdata <- SingleCellExperiment::colData(inSCE)

+  expres <- SummarizedExperiment::assay(inSCE, useAssay)

+  fdata <- SingleCellExperiment::rowData(inSCE)

+  SCENew <- MAST::FromMatrix(expres, pdata, fdata)

+

+  #Caculate CDR for zlm model

+  SummarizedExperiment::colData(SCENew)$cngeneson <-

+    scale(colSums(SummarizedExperiment::assay(SCENew) > 0))

+

+  if (useThresh){

+    SCENew <- SCENew[which(MAST::freq(SCENew) > 0), ]

+    invisible(utils::capture.output(thresh <- MAST::thresholdSCRNACountMatrix(

+      SummarizedExperiment::assay(SCENew), nbins = 20, min_per_bin = 30)))

+    SummarizedExperiment::assays(SCENew) <-

+      list(thresh = thresh$counts_threshold,

+           tpm = SummarizedExperiment::assay(SCENew))

+  }

+

+  # filter based on frequency of expression across samples

+  if (sum(MAST::freq(SCENew) > freqExpressed) <= 1){

+    stop("Not enough genes pass frequency expressed filter of 1")

+  }

+  SCENewSample <- SCENew[which(MAST::freq(SCENew) > freqExpressed), ]

+

+  # if the condition of interest is numeric, to change it to a factor

+  if (is.numeric(SummarizedExperiment::colData(SCENewSample)[, condition])){

+    SummarizedExperiment::colData(SCENewSample)[, condition] <-

+      as.factor(SummarizedExperiment::colData(SCENewSample)[, condition])

+  }

+  #Check for NAs, if true throw error

+  if (any(is.na(SingleCellExperiment::colData(inSCE)[, condition]))){

+     stop("Annotation data has NA values. Filter them to continue.")

+  }

+  # >2 levels in the condition

+  if (!is.null(interest.level) &

+      length(unique(SingleCellExperiment::colData(inSCE)[, condition])) > 2){

+    levels(SummarizedExperiment::colData(SCENewSample)[, condition]) <-

+      c(levels(SummarizedExperiment::colData(SCENewSample)[, condition]),

+        paste0("no_", interest.level))

+    SummarizedExperiment::colData(SCENewSample)[, condition][

+      SummarizedExperiment::colData(SCENewSample)[, condition] !=

+        interest.level] <- paste0("no_", interest.level)

+    SummarizedExperiment::colData(SCENewSample)[, condition] <-

+      droplevels(as.factor(

+        SummarizedExperiment::colData(SCENewSample)[, condition]))

+

+    hurdle1 <- MAST::zlm(stats::as.formula(paste0("~", condition,

+                                                  "+cngeneson")), SCENewSample)

+    summaryh1 <- MAST::summary(hurdle1, doLRT = paste0(condition, "no_",

+                                                       interest.level))

+

+    summaryDT <- summaryh1[["datatable"]]

+

+    fcHurdle <- merge(

+      summaryDT[summaryDT$contrast == paste0(condition, "no_", interest.level) &

+                  summaryDT$component == "H", c("primerid", "Pr(>Chisq)")],

+      summaryDT[summaryDT$contrast == paste0(condition, "no_", interest.level) &

+                  summaryDT$component == "logFC", c("primerid", "coef", "ci.hi",

+                                                    "ci.lo")]

+    )

+  } else {

+    SummarizedExperiment::colData(SCENewSample)[, condition] <-

+      droplevels(as.factor(

+        SummarizedExperiment::colData(SCENewSample)[, condition]))

+    level.cond  <- levels(

+      SummarizedExperiment::colData(SCENewSample)[, condition])

+

+    hurdle1 <- MAST::zlm(stats::as.formula(paste0("~", condition,

+                                                  "+cngeneson")), SCENewSample)

+    summaryh1 <- MAST::summary(hurdle1, doLRT = paste0(condition,

+                                                       level.cond[2]))

+

+    summaryDT <- summaryh1[["datatable"]]

+

+    fcHurdle <- merge(

+      summaryDT[summaryDT$contrast == paste0(condition, level.cond[2]) &

+                  summaryDT$component == "H", c("primerid", "Pr(>Chisq)")],

+      summaryDT[summaryDT$contrast == paste0(condition, level.cond[2]) &

+                  summaryDT$component == "logFC", c("primerid", "coef", "ci.hi",

+                                                    "ci.lo")]

+    )

+  }

+

+  # Use p-value correction method, here we use fdr

+  fcHurdle$fdr <- stats::p.adjust(fcHurdle$"Pr(>Chisq)", "fdr")

+

+  # Filter the data again by the adjusted pvalue and coef

+  fcHurdleSig <-  fcHurdle[fcHurdle$fdr < p.value & abs(fcHurdle$coef) >

+                             fcThreshold & !is.nan(fcHurdle$coef), ]

+  colnames(fcHurdleSig)[1] <- "Gene"

+

+  data.table::setorder(fcHurdleSig, "fdr")

+

+  return(fcHurdleSig)

+}

@@ -79,7 +79,7 @@

   ## Get merged rowData

   by.r <- unique(c('rownames', by.r))

-  unionFe <- Reduce(function(r1, r2) merge(r1, r2, by=by.r, all=T), feList)

+  unionFe <- Reduce(function(r1, r2) merge(r1, r2, by=by.r, all=TRUE), feList)

   allGenes <- unique(unlist(lapply(feList, rownames)))

   ## rowData

@@ -104,7 +104,7 @@

   })

   by.c <- unique(c("rownames", by.c))

-  unionCb <- Reduce(function(c1, c2) merge(c1, c2, by=by.c, all=T), cbList)

+  unionCb <- Reduce(function(c1, c2) merge(c1, c2, by=by.c, all=TRUE), cbList)

   rownames(unionCb) <- unionCb[['rownames']]

   newCbList <- list()

   for (i in seq_along(sceList)) {

@@ -234,7 +234,7 @@ combineSCE <- function(sceList, by.r, by.c, combined){

   assayList <- .mergeAssaySCE(sceList) 

   New_SCE <- list()

-  for (i in 1:length(sceList)) {

+  for (i in seq(length(sceList))) {

     ## create new sce

     New_SCE[[i]] <- .constructSCE(matrices = assayList[[i]], features = newFeList,

                                   barcodes = newCbList[[i]], 

@@ -249,4 +249,4 @@ combineSCE <- function(sceList, by.r, by.c, combined){

     return(sce)

   }

   return(New_SCE)

-}

\ No newline at end of file

+}

@@ -1,7 +1,7 @@

 #' Example Single Cell RNA-Seq data in SingleCellExperiment Object, GSE60361

 #' subset

 #'

-#' A subset of 30 samples from a single cell RNA-Seq experiment from Zeisel, et

+#' A subset of 30 cells from a single cell RNA-Seq experiment from Zeisel, et

 #' al. Science 2015. The data was produced from cells from the mouse

 #' somatosensory cortex (S1) and hippocampus (CA1). 15 of the cells were

 #' identified as oligodendrocytes and 15 of the cell were identified as

@@ -16,12 +16,12 @@

 #' data("mouseBrainSubsetSCE")

 "mouseBrainSubsetSCE"

-#' Example Single Cell RNA-Seq data in SingleCellExperiment Object, 

+#' Example Single Cell RNA-Seq data in SingleCellExperiment Object,

 #' subset of 10x public dataset

 #' https://support.10xgenomics.com/single-cell-gene-expression/datasets/2.1.0/pbmc4k

 #' A subset of 390 barcodes and top 200 genes were included in this example.

 #' Within 390 barcodes, 195 barcodes are empty droplet, 150 barcodes are cell barcode

-#' and 45 barcodes are doublets predicted by scrublet and doubletFinder package. 

+#' and 45 barcodes are doublets predicted by scrublet and doubletFinder package.

 #' This example only serves as a proof of concept and a tutoriol on how to

 #' run the functions in this package. The results should not be

 #' used for drawing scientific conclusions.

@@ -40,38 +40,31 @@

 #' Two batches of pancreas scRNAseq dataset are combined with their original

 #' counts. Cell types and batches are annotated in `colData(sceBatches)`.

 #' Two batches came from Wang, et al., 2016, annotated as `'w'`; and Xin, et

-#' al., 2016, annotated as `'x'`. Two common cell types, `'alpha'` and 

-#' `'beta'`, that could be found in both original studies with relatively 

+#' al., 2016, annotated as `'x'`. Two common cell types, `'alpha'` and

+#' `'beta'`, that could be found in both original studies with relatively

 #' large population were kept for cleaner demonstration.

-#'

-#' @name sceBatches

-#' @docType data

-#' @format SingleCellExperiment

-#' @source DOI: 10.2337/db16-0405 and 10.1016/j.cmet.2016.08.018

-#' @keywords datasets

-#' @examples

 #' data('sceBatches')

 "sceBatches"

 #' Stably Expressed Gene (SEG) list obect, with SEG sets for human and mouse.

-#' 

+#'

 #' The two gene sets came from dataset called `segList` of package `scMerge`.

 #' @name SEG

 #' @docType data

-#' @format list, with two entries `"human"` and `"mouse"`, each is a charactor

-#' array.

-#' @source `data('segList', package='scMerge')``

+#' @format list, with two entries \code{"human"} and \code{"mouse"}, each is a

+#' charactor vector.

+#' @source \code{data('segList', package='scMerge')}

 #' @keywords datasets

-#' @examples 

+#' @examples

 #' data('SEG')

 #' humanSEG <- SEG$human

 "SEG"

 #' MSigDB gene get Cctegory table

-#' 

-#' A table of gene set categories that can be download from MSigDB. The 

+#'

+#' A table of gene set categories that can be download from MSigDB. The

 #' categories and descriptions can be found here:

-#' https://www.gsea-msigdb.org/gsea/msigdb/collections.jsp. The IDs in the 

+#' https://www.gsea-msigdb.org/gsea/msigdb/collections.jsp. The IDs in the

 #' first column can be used to retrieve the gene sets for these categories

 #' using the \link{importGeneSetsFromMSigDB} function.

@@ -1,427 +1,666 @@

-## Original code from DoubletFinder package

-## (https://github.com/chris-mcginnis-ucsf/DoubletFinder)

-

-.parallel_paramSweep <- function(n, n.real.cells, real.cells, pK, pN, data,

-                                   orig.commands, PCs, sct, verbose, seed) {

-    sweep.res.list <- list()

-    list.ind <- 0

-

-    ## Make merged real-artifical data

-    if (verbose) {

-        print(paste("Creating artificial doublets for pN = ",

-                    pN[n] * 100, "%", sep = ""))

-    }

-    n_doublets <- ceiling(n.real.cells / (1 - pN[n]) - n.real.cells)

-    real.cells1 <- sample(real.cells, n_doublets, replace = TRUE)

-    real.cells2 <- sample(real.cells, n_doublets, replace = TRUE)

-    doublets <- (data[, real.cells1] + data[, real.cells2]) / 2

-    colnames(doublets) <- paste("X", 1:n_doublets, sep = "")

-    data_wdoublets <- cbind(data, doublets)

-

-    ## Pre-process Seurat object

-    if (sct == FALSE) {

-        if (verbose) {

-            print("Creating Seurat object...")

-        }

-        seu_wdoublets <- Seurat::CreateSeuratObject(counts = data_wdoublets)

-

-        if (verbose) {

-            print("Normalizing Seurat object...")

-        }

-        seu_wdoublets <- Seurat::NormalizeData(seu_wdoublets,

-                                       normalization.method = orig.commands$NormalizeData.RNA@params$normalization.method,

-                                       scale.factor = orig.commands$NormalizeData.RNA@params$scale.factor,

-                                       margin = orig.commands$NormalizeData.RNA@params$margin,

-                                       verbose = verbose

-        )

-

-        if (verbose) {

-            print("Finding variable genes...")

-        }

-        seu_wdoublets <- Seurat::FindVariableFeatures(seu_wdoublets,

-                                              selection.method = orig.commands$FindVariableFeatures.RNA$selection.method,

-                                              loess.span = orig.commands$FindVariableFeatures.RNA$loess.span,

-                                              clip.max = orig.commands$FindVariableFeatures.RNA$clip.max,

-                                              mean.function = orig.commands$FindVariableFeatures.RNA$mean.function,

-                                              dispersion.function = orig.commands$FindVariableFeatures.RNA$dispersion.function,

-                                              num.bin = orig.commands$FindVariableFeatures.RNA$num.bin,

-                                              binning.method = orig.commands$FindVariableFeatures.RNA$binning.method,

-                                              nfeatures = orig.commands$FindVariableFeatures.RNA$nfeatures,

-                                              mean.cutoff = orig.commands$FindVariableFeatures.RNA$mean.cutoff,

-                                              dispersion.cutoff = orig.commands$FindVariableFeatures.RNA$dispersion.cutoff,

-                                              verbose = verbose

-        )

-

-        if (verbose) {

-            print("Scaling data...")

-        }

-        seu_wdoublets <- Seurat::ScaleData(

-            seu_wdoublets,

-            features = orig.commands$ScaleData.RNA$features,

-            model.use = orig.commands$ScaleData.RNA$model.use,

-            do.scale = orig.commands$ScaleData.RNA$do.scale,

-            do.center = orig.commands$ScaleData.RNA$do.center,

-            scale.max = orig.commands$ScaleData.RNA$scale.max,

-            block.size = orig.commands$ScaleData.RNA$block.size,

-            min.cells.to.block = orig.commands$ScaleData.RNA$min.cells.to.block,

-            verbose = verbose

-        )

-

-        if (verbose) {

-            print("Running PCA...")

-        }

-        seu_wdoublets <- Seurat::RunPCA(

-            seu_wdoublets,

-            features = orig.commands$ScaleData.RNA$features,

-            npcs = length(PCs),

-            rev.pca = orig.commands$RunPCA.RNA$rev.pca,

-            weight.by.var = orig.commands$RunPCA.RNA$weight.by.var,

-            seed.use = seed,

-            verbose = FALSE

-        )

-    }

-

-    if (sct == TRUE) {

-        if (verbose) {

-            print("Creating Seurat object...")

-        }

-        seu_wdoublets <- Seurat::CreateSeuratObject(counts = data_wdoublets)

-

-        if (verbose) {

-            print("Running SCTransform...")

-        }

-        seu_wdoublets <- Seurat::SCTransform(seu_wdoublets)

-

-        if (verbose) {

-            print("Running PCA...")

-        }

-        seu_wdoublets <- Seurat::RunPCA(seu_wdoublets,

-                                npcs = length(PCs), seed.use = seed,

-                                verbose = verbose

-        )

-    }

-

-    ## Compute PC distance matrix

-    if (verbose) {

-        print("Calculating PC distance matrix...")

-    }

-    nCells <- nrow(seu_wdoublets@meta.data)

-    pca.coord <- seu_wdoublets@reductions$pca@cell.embeddings[, PCs]

-    seu_wdoublets <- NULL

-    gc()

-    dist.mat <- fields::rdist(pca.coord)[, 1:n.real.cells]

-

-    ## Pre-order PC distance matrix prior to iterating across pK

-    ## for pANN computations

-    if (verbose) {

-        print("Defining neighborhoods...")

-    }

-

-    for (i in 1:n.real.cells) {

-        dist.mat[, i] <- order(dist.mat[, i])

-    }

-

-    ## Trim PC distance matrix for faster manipulations

-    ind <- round(nCells * max(pK)) + 5

-    dist.mat <- dist.mat[1:ind, ]

-

-    ## Compute pANN across pK sweep

-

-    if (verbose) {

-        print("Computing pANN across all pK...")

-    }

-    for (k in 1:length(pK)) {

-        if (verbose) {

-            print(paste("pK = ", pK[k], "...", sep = ""))

-        }

-        pk.temp <- round(nCells * pK[k])

-        pANN <- as.data.frame(matrix(0L, nrow = n.real.cells, ncol = 1))

-        colnames(pANN) <- "pANN"

-        rownames(pANN) <- real.cells

-        list.ind <- list.ind + 1

-

-        for (i in 1:n.real.cells) {

-            neighbors <- dist.mat[2:(pk.temp + 1), i]

-            pANN$pANN[i] <- length(which(neighbors > n.real.cells)) / pk.temp

-        }

-

-        sweep.res.list[[list.ind]] <- pANN

-    }

-

-    return(sweep.res.list)

-}

-

-.paramSweep <- function(seu, PCs = 1:10, sct = FALSE,

-                          verbose = FALSE, num.cores, seed) {

-    ## Set pN-pK param sweep ranges

-    pK <- c(0.0005, 0.001, 0.005, seq(0.01, 0.3, by = 0.01))

-    pN <- seq(0.05, 0.3, by = 0.05)

-    ## Remove pK values with too few cells

-    min.cells <- round(nrow(seu@meta.data) / (1 - 0.05) - nrow(seu@meta.data))

-    pK.test <- round(pK * min.cells)

-    pK <- pK[which(pK.test >= 1)]

-

-    ## Extract pre-processing parameters from original data analysis workflow

-    orig.commands <- seu@commands

-

-    ## Down-sample cells to 10000 (when applicable) for computational effiency

-    if (nrow(seu@meta.data) > 10000) {

-        real.cells <- rownames(seu@meta.data)[sample(1:nrow(seu@meta.data),

-                                                     10000, replace = FALSE)]

-        data <- seu@assays$RNA@counts[, real.cells]

-        n.real.cells <- ncol(data)

-    }

-

-    if (nrow(seu@meta.data) <= 10000) {

-        real.cells <- rownames(seu@meta.data)

-        data <- seu@assays$RNA@counts

-        n.real.cells <- ncol(data)

-    }

-    ## Iterate through pN, computing pANN vectors at varying pK

-    # no_cores <- detectCores()-1

-    if (is.null(num.cores)) {

-        num.cores <- 1

-    }

-

-    if (num.cores > 1) {

-        cl <- parallel::makeCluster(num.cores)

-

-        output2 <- withr::with_seed(

-            seed,

-            parallel::mclapply(as.list(1:length(pN)),

-                     FUN = .parallel_paramSweep,

-                     n.real.cells,

-                     real.cells,

-                     pK,

-                     pN,

-                     data,

-                     orig.commands,

-                     PCs,

-                     sct, mc.cores = num.cores,

-                     verbose = verbose,

-                     seed = seed,

-                     mc.set.seed = FALSE

-            )

-        )

-        parallel::stopCluster(cl)

-    } else {

-        output2 <- lapply(as.list(1:length(pN)),

-                          FUN = .parallel_paramSweep,

-                          n.real.cells,

-                          real.cells,

-                          pK,

-                          pN,

-                          data,

-                          orig.commands,

-                          PCs,

-                          sct,

-                          seed = seed,

-                          verbose = verbose

-        )

-    }

-

-    ## Write parallelized output into list

-    sweep.res.list <- list()

-    list.ind <- 0

-    for (i in 1:length(output2)) {

-        for (j in 1:length(output2[[i]])) {

-            list.ind <- list.ind + 1

-            sweep.res.list[[list.ind]] <- output2[[i]][[j]]

-        }

-    }

-    ## Assign names to list of results

-    name.vec <- NULL

-    for (j in 1:length(pN)) {

-        name.vec <- c(name.vec, paste("pN", pN[j], "pK", pK, sep = "_"))

-    }

-    names(sweep.res.list) <- name.vec

-    return(sweep.res.list)

-

-}

-

-.runDoubletFinder <- function(counts, seuratPcs, seuratRes, formationRate,

-                              seuratNfeatures, verbose = FALSE,

-                              nCores = NULL, seed = 12345) {

-

-  ## Convert to sparse matrix if not already in that format

-  counts <- .convertToMatrix(counts)

-  colnames(counts) <- gsub("_", "-", colnames(counts))

-

-  seurat <- suppressWarnings(Seurat::CreateSeuratObject(

-    counts = counts,

-    project = "seurat", min.features = 0

-  ))

-  seurat <- Seurat::NormalizeData(

-    object = seurat,

-    normalization.method = "LogNormalize", scale.factor = 10000,

-    verbose = verbose

-  )

-  seurat <- Seurat::FindVariableFeatures(seurat,

-    selection.method = "vst",

-    nfeatures = seuratNfeatures,

-    verbose = verbose

-  )

-

-  allGenes <- rownames(seurat)

-  seurat <- Seurat::ScaleData(seurat, features = allGenes, verbose = verbose)

-

-  numPc <- min(ncol(seurat@assays$RNA@scale.data) - 1, 50)

-  seurat <- Seurat::RunPCA(seurat,

-    features =

-      Seurat::VariableFeatures(object = seurat),

-    npcs = numPc, verbose = verbose, seed.use = seed

-  )

-  seurat <- Seurat::FindNeighbors(seurat, dims = seuratPcs, verbose = verbose)

-  seurat <- Seurat::FindClusters(seurat,

-    resolution = seuratRes,

-    verbose = verbose,

-    random.seed = seed

-  )

-  invisible(sweepResListSeurat <- .paramSweep(seurat,

-    PCs = seuratPcs,

-    sct = FALSE,

-    num.cores = nCores,

-    verbose = verbose,

-    seed = seed

-  ))

-  sweepStatsSeurat <- DoubletFinder::summarizeSweep(sweepResListSeurat,

-    GT = FALSE

-  )

-  bcmvnSeurat <- DoubletFinder::find.pK(sweepStatsSeurat, verbose = verbose)

-  pkOptimal <- as.numeric(as.matrix(bcmvnSeurat$pK[

-    which.max(bcmvnSeurat$MeanBC)

-  ]))

-  annotations <- seurat@meta.data$seurat_clusters

-  homotypicProp <- DoubletFinder::modelHomotypic(annotations)

-  nExpPoi <- round(formationRate * ncol(seurat@assays$RNA))

-  seurat <- DoubletFinder::doubletFinder_v3(seurat,

-    PCs = seuratPcs,

-    pN = 0.25,

-    pK = pkOptimal,

-    nExp = nExpPoi,

-    reuse.pANN = FALSE,

-    sct = FALSE,

-    verbose = FALSE

-  )

-  names(seurat@meta.data)[6] <- "doubletFinderAnnScore"

-  names(seurat@meta.data)[7] <- "doubletFinderLabel"

-  return(seurat)

-}

-

-#' @title Generates a doublet score for each cell via doubletFinder

-#' @description Uses doubletFinder to determine cells within the dataset

-#'  suspected to be doublets.

-#' @param inSCE Input SingleCellExperiment object. Must contain a counts matrix

-#' @param useAssay  Indicate which assay to use. Default "counts".

-#' @param sample Numeric vector. Each cell will be assigned a sample number.

-#' @param seed Seed for the random number generator. Default 12345.

-#' @param seuratNfeatures Integer. Number of highly variable genes to use.

-#'  Default 2000.

-#' @param seuratPcs Numeric vector. The PCs used in seurat function to

-#'   determine number of clusters. Default 1:15.

-#' @param seuratRes Numeric vector. The resolution parameter used in seurat,

-#'  which adjusts the number of clusters determined via the algorithm.

-#'  Default 1.5.

-#' @param formationRate Doublet formation rate used within algorithm.

-#'  Default 0.075.

-#' @param nCores Number of cores used for running the function.

-#' @param verbose Boolean. Wheter to print messages from Seurat and

-#'  DoubletFinder. Default FALSE.

-#' @return SingleCellExperiment object containing the

-#'  'doublet_finder_doublet_score'.

-#' @examples

-#' \donttest{

-#' data(scExample, package = "singleCellTK")

-#' sce <- sce[, colData(sce)$type != 'EmptyDroplet']

-#' sce <- runDoubletFinder(sce)

-#' }

-#' @export

-#' @importFrom SummarizedExperiment colData colData<-

-#' @importFrom SingleCellExperiment reducedDim<-

-runDoubletFinder <- function(inSCE,

-                             useAssay = "counts",

-                             sample = NULL,

-                             seed = 12345,

-                             seuratNfeatures = 2000,

-                             seuratPcs = 1:15,

-                             seuratRes = 1.5,

-                             formationRate = 0.075,

-                             nCores = NULL,

-                             verbose = FALSE) {

-

-  # argsList <- as.list(formals(fun = sys.function(sys.parent()), envir = parent.frame()))

-  argsList <- mget(names(formals()), sys.frame(sys.nframe()))

-

-  if (!is.null(sample)) {

-    if (length(sample) != ncol(inSCE)) {

-      stop("'sample' must be the same length as the number of columns in 'inSCE'")

-    }

-  } else {

-    sample <- rep(1, ncol(inSCE))

-  }

-

-  message(paste0(date(), " ... Running 'doubletFinder'"))

-

-  doubletScore <- rep(NA, ncol(inSCE))

-  doubletLabel <- rep(NA, ncol(inSCE))

-  allSampleNumbers <- sort(unique(sample))

-

-

-  for (res in seuratRes) {

-    output <- S4Vectors::DataFrame(

-      row.names = colnames(inSCE),

-      doubletFinder_doublet_score = numeric(ncol(inSCE)),

-      doubletFinder_doublet_label = numeric(ncol(inSCE))

-    )

-    umapDims <- matrix(ncol = 2,

-                       nrow = ncol(inSCE))

-    rownames(umapDims) = colnames(inSCE)

-    colnames(umapDims) = c("UMAP_1", "UMAP_2")

-

-    ## Loop through each sample and run doubletFinder

-    samples <- unique(sample)