@@ -99,7 +99,7 @@ Imports:

     TENxPBMCData,

     yaml,

     rmarkdown

-RoxygenNote: 7.1.0

+RoxygenNote: 7.1.1

 Suggests:

     testthat,

     Rsubread,

@@ -1,40 +1,41 @@

-#' Apply the mutual nearest neighbors (MNN) batch effect correction method to 

+#' Apply the mutual nearest neighbors (MNN) batch effect correction method to

 #' SingleCellExperiment object

-#' 

-#' MNN is designed for batch correction of single-cell RNA-seq data where the 

-#' batches are partially confounded with biological conditions of interest. It 

-#' does so by identifying pairs of MNN in the high-dimensional log-expression 

-#' space. For each MNN pair, a pairwise correction vector is computed by 

+#'

+#' MNN is designed for batch correction of single-cell RNA-seq data where the

+#' batches are partially confounded with biological conditions of interest. It

+#' does so by identifying pairs of MNN in the high-dimensional log-expression

+#' space. For each MNN pair, a pairwise correction vector is computed by

 #' applying a Gaussian smoothing kernel with bandwidth `sigma`.

-#' @param inSCE SingleCellExperiment object. An object that stores your dataset

-#' and analysis procedures.

-#' @param useAssay character, default `"logcounts"`. A string indicating the name 

-#' of the assay requiring batch correction in "inSCE", should exist in 

-#' `assayNames(inSCE)`.

-#' @param batch character, default `"batch"`. A string indicating the 

-#' field of `colData(inSCE)` that defines different batches.

-#' @param assayName character, default `"MNN"`. The name for the corrected 

-#' full-sized expression matrix.

-#' @param k integer, default `20`. Specifies the number of nearest neighbours to 

-#' consider when defining MNN pairs. This should be interpreted as the minimum 

-#' frequency of each cell type or state in each batch. Larger values will 

-#' improve the precision of the correction by increasing the number of MNN 

+#' @param inSCE \linkS4class{SingleCellExperiment} inherited object. Required.

+#' @param useAssay A single character indicating the name of the assay requiring

+#' batch correction. Default \code{"logcounts"}.

+#' @param batch A single character indicating a field in

+#' \code{\link[SummarizedExperiment]{colData}} that annotates the batches.

+#' Default \code{"batch"}.

+#' @param k An integer. Specifies the number of nearest neighbours to

+#' consider when defining MNN pairs. This should be interpreted as the minimum

+#' frequency of each cell type or state in each batch. Larger values will

+#' improve the precision of the correction by increasing the number of MNN

 #' pairs, at the cost of reducing accuracy by allowing MNN pairs to form between

-#' cells of different type.

-#' @param sigma Numeric, default `0.1`. Specifies how much information is 

-#' shared between MNN pairs when computing the batch effect. Larger values will 

-#' share more information, approaching a global correction for all cells in the 

-#' same batch. Smaller values allow the correction to vary across cell types, 

-#' which may be more accurate but comes at the cost of precision. 

-#' @return SingleCellExperiment object with `reducedDim(inSCE, reducedDimName)` 

-#' updated with corrected low-dimentional representation.

+#' cells of different type. Default \code{20L}.

+#' @param sigma A Numeric scalar. Specifies how much information is

+#' shared between MNN pairs when computing the batch effect. Larger values will

+#' share more information, approaching a global correction for all cells in the

+#' same batch. Smaller values allow the correction to vary across cell types,

+#' which may be more accurate but comes at the cost of precision. Default

+#' \code{0.1}.

+#' @param assayName A single characeter. The name for the corrected assay. Will

+#' be saved to \code{\link[SummarizedExperiment]{assay}}. Default

+#' \code{"MNN"}.

+#' @return The input \linkS4class{SingleCellExperiment} object with

+#' \code{assay(inSCE, assayName)} updated.

 #' @export

-#' @references Lun ATL, et al., 2016 & 2018 

-#' @examples  

+#' @references Lun ATL, et al., 2016 & 2018

+#' @examples

 #' data('sceBatches', package = 'singleCellTK')

 #' sceCorr <- runMNNCorrect(sceBatches)

-runMNNCorrect <- function(inSCE, useAssay = 'logcounts', batch = 'batch', 

-                          assayName = 'MNN', k = 20, sigma = 0.1){

+runMNNCorrect <- function(inSCE, useAssay = 'logcounts', batch = 'batch',

+                          assayName = 'MNN', k = 20L, sigma = 0.1){

     ## Input check

     if(!inherits(inSCE, "SingleCellExperiment")){

         stop("\"inSCE\" should be a SingleCellExperiment Object.")

@@ -46,11 +47,12 @@ runMNNCorrect <- function(inSCE, useAssay = 'logcounts', batch = 'batch',

         stop(paste("\"batch name:", batch, "not found."))

     }

     assayName <- gsub(' ', '_', assayName)

-    

+    k <- as.integer(k)

+

     ## Run algorithm

     batchCol <- SummarizedExperiment::colData(inSCE)[[batch]]

     batchFactor <- as.factor(batchCol)

-    mnnSCE <- batchelor::mnnCorrect(inSCE, batch = batchFactor, 

+    mnnSCE <- batchelor::mnnCorrect(inSCE, batch = batchFactor,

                                     k = k, sigma = sigma)

     corrected <- SummarizedExperiment::assay(mnnSCE, 'corrected')

     SummarizedExperiment::assay(inSCE, assayName) <- corrected

@@ -4,23 +4,23 @@

 #' SCANORAMA is analogous to computer vision algorithms for panorama stitching

 #' that identify images with overlapping content and merge these into a larger

 #' panorama.

-#' @param inSCE SingleCellExperiment object. An object that stores your dataset

-#' and analysis procedures.

-#' @param useAssay character, default `"logcounts"`. A string indicating the name

-#' of the assay requiring batch correction in "inSCE", should exist in

-#' `assayNames(inSCE)`.

-#' @param batch character, default `"batch"`. A string indicating the

-#' field of `colData(inSCE)` that defines different batches.

-#' @param assayName character, default `"SCANORAMA"`. The name for the

-#' corrected full-sized expression matrix.

-#' @param SIGMA numeric, default `15`. Algorithmic parameter, correction

-#' smoothing parameter on Gaussian kernel.

-#' @param ALPHA numeric, default `0.1`. Algorithmic parameter, alignment score

-#' minimum cutoff.

-#' @param KNN integer, default `20L`. Algorithmic parameter, number of nearest

-#' neighbors to use for matching.

-#' @return SingleCellExperiment object with `assay(inSCE, assayName)` updated

-#' with corrected full-sized expression matrix.

+#' @param inSCE \linkS4class{SingleCellExperiment} inherited object. Required.

+#' @param useAssay A single character indicating the name of the assay requiring

+#' batch correction. Default \code{"logcounts"}.

+#' @param batch A single character indicating a field in

+#' \code{\link[SummarizedExperiment]{colData}} that annotates the batches.

+#' Default \code{"batch"}.

+#' @param SIGMA A numeric scalar. Algorithmic parameter, correction smoothing

+#' parameter on Gaussian kernel. Default \code{15}.

+#' @param ALPHA A numeric scalar. Algorithmic parameter, alignment score

+#' minimum cutoff. Default \code{0.1}.

+#' @param KNN An integer. Algorithmic parameter, number of nearest neighbors to

+#' use for matching. Default \code{20L}.

+#' @param assayName A single characeter. The name for the corrected assay. Will

+#' be saved to \code{\link[SummarizedExperiment]{assay}}. Default

+#' \code{"SCANORAMA"}.

+#' @return The input \linkS4class{SingleCellExperiment} object with

+#' \code{assay(inSCE, assayName)} updated.

 #' @export

 #' @references Brian Hie et al, 2019

 #' @examples

@@ -29,8 +29,8 @@

 #' sceCorr <- runSCANORAMA(sceBatches)

 #' }

 runSCANORAMA <- function(inSCE, useAssay = 'logcounts', batch = 'batch',

-                         assayName = 'SCANORAMA', SIGMA = 15, ALPHA = 0.1,

-                         KNN = 20L){

+                         SIGMA = 15, ALPHA = 0.1, KNN = 20L,

+                         assayName = 'SCANORAMA'){

     ## Input check

     if(!inherits(inSCE, "SingleCellExperiment")){

         stop("\"inSCE\" should be a SingleCellExperiment Object.")

@@ -1,35 +1,35 @@

 #' Apply scGen batch effect correction method to SingleCellExperiment object

-#' 

-#' scGen is a generative model to predict single-cell perturbation response 

-#' across cell types, studies and species. It works by combining variational 

+#'

+#' scGen is a generative model to predict single-cell perturbation response

+#' across cell types, studies and species. It works by combining variational

 #' autoencoders and latent space vector arithmetics for high-dimensional single-

 #' cell gene expression data.

-#' 

-#' Result does not look fine for now. Time consuming also even it allocates 32 

-#' cores.

-#' @param inSCE SingleCellExperiment object. An object that stores your dataset

-#' and analysis procedures.

-#' @param useAssay character, default `"logcounts"`. A string indicating the name 

-#' of the assay requiring batch correction in "inSCE", should exist in 

-#' `assayNames(inSCE)`.

-#' @param batch character, default `"batch"`. A string indicating the field 

-#' of `colData(inSCE)` that defines different batches.

-#' @param cellType character, default `"cell_type"`. A string indicating the 

-#' field of `colData(inSCE)` that defines different cell types.

-#' @param assayName character, default `"SCGEN"`. The name for the corrected 

-#' full-sized expression matrix.

-#' @param nEpochs integer, default `100L`. Algorithmic parameter, number of 

-#' epochs to iterate and optimize network weights. 

+#' @param inSCE \linkS4class{SingleCellExperiment} inherited object. Required.

+#' @param useAssay A single character indicating the name of the assay requiring

+#' batch correction. Default \code{"logcounts"}.

+#' @param batch A single character indicating a field in

+#' \code{\link[SummarizedExperiment]{colData}} that annotates the batches.

+#' Default \code{"batch"}.

+#' @param cellType A single character. A string indicating a field in

+#' \code{colData(inSCE)} that defines different cell types. Default

+#' \code{'cell_type'}.

+#' @param nEpochs An integer. Algorithmic parameter, the number of epochs to

+#' iterate and optimize network weights. Default \code{50L}.

+#' @param assayName A single characeter. The name for the corrected assay. Will

+#' be saved to \code{\link[SummarizedExperiment]{assay}}. Default

+#' \code{"SCGEN"}.

+#' @return The input \linkS4class{SingleCellExperiment} object with

+#' \code{assay(inSCE, assayName)} updated.

 #' @export

 #' @references Lotfollahi, Mohammad et al., 2019

-#' @examples 

+#' @examples

 #' \dontrun{

 #' data('sceBatches', package = 'singleCellTK')

 #' sceCorr <- runSCGEN(sceBatches)

 #' }

-runSCGEN <- function(inSCE, useAssay = 'logcounts', batch = 'batch', 

-                     cellType = "cell_type", assayName = 'SCGEN', 

-                     nEpochs = 50L){

+runSCGEN <- function(inSCE, useAssay = 'logcounts', batch = 'batch',

+                     cellType = "cell_type", nEpochs = 50L,

+                     assayName = 'SCGEN'){

     ## Input check

     if(!inherits(inSCE, "SingleCellExperiment")){

         stop("\"inSCE\" should be a SingleCellExperiment Object.")

@@ -59,7 +59,7 @@ runSCGEN <- function(inSCE, useAssay = 'logcounts', batch = 'batch',

     adata <- .sce2adata(inSCE, useAssay = useAssay)

     network = scgen$VAEArith(x_dimension = adata$n_vars)

     network$train(train_data = adata, n_epochs = nEpochs)

-    corrAdata <- scgen$batch_removal(network, adata, batch_key = batch, 

+    corrAdata <- scgen$batch_removal(network, adata, batch_key = batch,

                                      cell_label_key = cellType)

     corrMat <- t(corrAdata$X)

     SummarizedExperiment::assay(inSCE, assayName) <- corrMat

@@ -1,40 +1,43 @@

 #' Apply scMerge batch effect correction method to SingleCellExperiment object

 #'

-#' The scMerge method leverages factor analysis, stably expressed genes (SEGs) 

-#' and (pseudo-) replicates to remove unwanted variations and merge multiple 

-#' scRNA-Seq data. 

-#' @param inSCE SingleCellExperiment object. An object that stores your dataset

-#' and analysis procedures.

-#' @param useAssay character, default `"logcounts"`. A string indicating the name 

-#' of the assay requiring batch correction in "inSCE", should exist in 

-#' `assayNames(inSCE)`.

-#' @param batch character, default `"batch"`. A string indicating the field 

-#' of `colData(inSCE)` that defines different batches.

-#' @param assayName character, default `"scMerge"`. The name for the corrected 

-#' full-sized expression matrix.

-#' @param kmeansK vector of int, default `NULL`. A vector indicating the 

-#' kmeans' K-value for each batch, in order to construct pseudo-replicates. The 

-#' length of `kmeansK` needs to be the same as the number of batches.

-#' @param cellType character, default `"cell_type"`. A string indicating the 

-#' field of `colData(inSCE)` that defines different cell types.

-#' @param seg array, default `NULL`. An array of gene names or indices that 

-#' specifies SEG (Stably Expressed Genes) set as negative control. Pre-defined 

-#' dataset with human and mouse SEG lists is available to user by running 

-#' `data('SEG')`.

-#' @param nCores integer, default `parallel::detectCores()`. The number of 

-#' cores of processors to allocate for the task. By default it takes all the 

-#' cores available to the user. 

-#' @return SingleCellExperiment object with `assay(inSCE, assayName)` updated 

-#' with corrected full-sized expression matrix.

+#' The scMerge method leverages factor analysis, stably expressed genes (SEGs)

+#' and (pseudo-) replicates to remove unwanted variations and merge multiple

+#' scRNA-Seq data.

+#' @param inSCE \linkS4class{SingleCellExperiment} inherited object. Required.

+#' @param useAssay A single character indicating the name of the assay requiring

+#' batch correction. Default \code{"logcounts"}.

+#' @param batch A single character indicating a field in

+#' \code{\link[SummarizedExperiment]{colData}} that annotates the batches.

+#' Default \code{"batch"}.

+#' @param kmeansK An integer vector. Indicating the kmeans' K-value for each

+#' batch (i.e. how many subclusters in each batch should exist), in order to

+#' construct pseudo-replicates. The length of code{kmeansK} needs to be the same

+#' as the number of batches. Default \code{NULL}, and this value will be

+#' auto-detected by default, depending on \code{cellType}.

+#' @param cellType A single character. A string indicating a field in

+#' \code{colData(inSCE)} that defines different cell types. Default

+#' \code{'cell_type'}.

+#' @param seg A vector of gene names or indices that specifies SEG (Stably

+#' Expressed Genes) set as negative control. Pre-defined dataset with human and

+#' mouse SEG lists is available to user by running \code{data('SEG')}. Default

+#' \code{NULL}, and this value will be auto-detected by default with

+#' \code{\link[scMerge]{scSEGIndex}}.

+#' @param nCores An integer. The number of cores of processors to allocate for

+#' the task. Default \code{1L}.

+#' @param assayName A single characeter. The name for the corrected assay. Will

+#' be saved to \code{\link[SummarizedExperiment]{assay}}. Default

+#' \code{"scMerge"}.

+#' @return The input \linkS4class{SingleCellExperiment} object with

+#' \code{assay(inSCE, assayName)} updated.

 #' @export

 #' @references Hoa, et al., 2020

-#' @examples 

+#' @examples

 #' data('sceBatches', package = 'singleCellTK')

 #' sceCorr <- runSCMerge(sceBatches)

-runSCMerge <- function(inSCE, useAssay = "logcounts", batch = 'batch', 

-                       assayName = "scMerge", seg = NULL, kmeansK = NULL, 

-                       cellType = 'cell_type', 

-                       nCores = 1){

+runSCMerge <- function(inSCE, useAssay = "logcounts", batch = 'batch',

+                       assayName = "scMerge", seg = NULL, kmeansK = NULL,

+                       cellType = 'cell_type',

+                       nCores = 1L){

     ## Input check

     if(!inherits(inSCE, "SingleCellExperiment")){

         stop("\"inSCE\" should be a SingleCellExperiment Object.")

@@ -55,12 +58,12 @@ runSCMerge <- function(inSCE, useAssay = "logcounts", batch = 'batch',

     nCores <- min(as.integer(nCores), parallel::detectCores())

     assayName <- gsub(' ', '_', assayName)

-    

+

     ## Run algorithm

     batchCol <- SummarizedExperiment::colData(inSCE)[[batch]]

     uniqBatch <- unique(batchCol)

-    

+

     # Infer parameters

     if(is.null(cellType)){

         cellTypeCol <- NULL

@@ -89,17 +92,17 @@ runSCMerge <- function(inSCE, useAssay = "logcounts", batch = 'batch',

     } else {

         ctl <- seg

     }

-    

+

     # scMerge automatically search for the column called "batch"...

     colDataNames <- names(SummarizedExperiment::colData(inSCE))

     names(SummarizedExperiment::colData(inSCE))[colDataNames == batch] <- 'batch'

     bpParam <- BiocParallel::MulticoreParam(workers = nCores)

     inSCE <- scMerge::scMerge(sce_combine = inSCE, exprs = useAssay,

-                              hvg_exprs = useAssay, 

-                              assay_name = assayName, 

-                              ctl = ctl, kmeansK = kmeansK, 

+                              hvg_exprs = useAssay,

+                              assay_name = assayName,

+                              ctl = ctl, kmeansK = kmeansK,

                               #marker_list = topVarGenesPerBatch,

-                              cell_type = cellTypeCol, 

+                              cell_type = cellTypeCol,

                               BPPARAM = bpParam)

     colDataNames <- names(SummarizedExperiment::colData(inSCE))

     names(SummarizedExperiment::colData(inSCE))[colDataNames == 'batch'] <- batch

@@ -9,22 +9,22 @@

 #' dataset, that is hypothesized to originate from the same cell state. These

 #' anchors are then used to harmonize the datasets, or transfer information

 #' from one dataset to another.

-#' @param inSCE SingleCellExperiment object. An object that stores your dataset

-#' and analysis procedures.

-#' @param useAssay character, default `"logcounts"`. A string indicating the name

-#' of the assay requiring batch correction in "inSCE", should exist in

-#' `assayNames(inSCE)`.

-#' @param batch character, default `"batch"`. A string indicating the

-#' field of `colData(inSCE)` that defines different batches.

-#' @param altExpName character, default `"Seurat3Int"`. The name for the

-#' corrected full-sized expression matrix. If the number of features returned

-#' is smaller the number of total feature, the returned matrix will be saved in

-#' `reducedDim(inSCE, assayName)`; if equal, `assay(inSCE, assayName)`.

-#' @param nAnchors integer, default `nrow(inSCE)`. The number of features to

-#' anchor, and also the final dimensionality of the integrated matrix. Thus

-#' default value turns to produce full-sized assay.

-#' @param verbose bool, default `TRUE`. Whether to show detail information of

-#' the process.

+#' @param inSCE \linkS4class{SingleCellExperiment} inherited object. Required.

+#' @param useAssay A single character indicating the name of the assay requiring

+#' batch correction. Default \code{"logcounts"}.

+#' @param batch A single character indicating a field in

+#' \code{\link[SummarizedExperiment]{colData}} that annotates the batches.

+#' Default \code{"batch"}.

+#' @param nAnchors An integer. The number of features to anchor. The final

+#' number of the corrected features depends on this value. Default

+#' \code{nrow(inSCE)}.

+#' @param verbose A logical scalar. Whether to show detail information of

+#' the process. Default \code{TRUE}.

+#' @param altExpName A single character. The name for the

+#' \code{\link[SingleCellExperiment]{altExp}} that stores the corrected assay.

+#' The name of this assay has the same name. Default \code{"Seurat3Int"}.

+#' @return The input \linkS4class{SingleCellExperiment} object with

+#' \code{altExp(inSCE, altExpName)} updated.

 #' @export

 #' @references Stuart et al. 2019

 #' @examples

@@ -1,69 +1,72 @@

-#' Apply ZINBWaVE Batch effect correction method to SingleCellExperiment object

-#'

-#' A general and flexible zero-inflated negative binomial model that can be

-#' used to provide a low-dimensional representations of scRNAseq data. The

-#' model accounts for zero inflation (dropouts), over-dispersion, and the count

-#' nature of the data. The model also accounts for the difference in library

-#' sizes and optionally for batch effects and/or other covariates.

-#' @param inSCE SingleCellExperiment object. An object that stores your dataset

-#' and analysis procedures.

-#' @param useAssay character, default `"logcounts"`. A string indicating the name

-#' of the assay requiring batch correction in "inSCE", should exist in

-#' `assayNames(inSCE)`.

-#' @param batch character, default `"batch"`. A string indicating the

-#' field of `colData(inSCE)` that defines different batches.

-#' @param reducedDimName character, default `"zinbwave"`. The name for the

-#' corrected low-dimensional representation.

-#' @param nHVG integer, default `1000`. Number of highly variable genes to use

-#' when fitting the model

-#' @param nComponents integer, default `50L`. Number of principle components or

-#' dimensionality to generate in the resulting reducedDim.

-#' @param nIter integer, default `10`. The max number of iterations to perform.

-#' @param epsilon integer, default `1000`. Algorithmic parameter, by default, the

-#' epsilon parameter is set to the number of genes. We empirically found that a

-#' high epsilon is often required to obtained a good low-level representation.

-#' @export

-#' @references Pollen, Alex A et al., 2014

-#' @examples

-#' \dontrun{

-#' data('sceBatches', package = 'singleCellTK')

-#' sceCorr <- runZINBWaVE(sceBatches, nIter=5)

-#' }

-runZINBWaVE <- function(inSCE, useAssay = 'logcounts', batch = 'batch',

-                        reducedDimName = 'zinbwave', nHVG = 1000,

-                        nComponents = 50, epsilon = 1000, nIter = 10){

-    #filterParams = NULL  <<< something told in tutorial but might be ignored

-    ## Input check

-    if(!inherits(inSCE, "SingleCellExperiment")){

-        stop("\"inSCE\" should be a SingleCellExperiment Object.")

-    }

-    if(!useAssay %in% SummarizedExperiment::assayNames(inSCE)) {

-        stop(paste("\"useAssay\" (assay) name: ", useAssay, " not found"))

-    }

-    if(!batch %in% names(SummarizedExperiment::colData(inSCE))){

-        stop(paste("\"batch name:", batch, "not found."))

-    }

-    reducedDimName <- gsub(' ', '_', reducedDimName)

-

-    # Run algorithm

-    tmpMatrix <- round(SummarizedExperiment::assay(inSCE, useAssay))

-    tmpSCE <- inSCE

-    SummarizedExperiment::assay(tmpSCE, useAssay) <- tmpMatrix

-

-    ##ZINBWaVE tutorial style of HVG selection

-    if(nHVG < nrow(inSCE)){

-        logAssay <- log1p(SummarizedExperiment::assay(tmpSCE, useAssay))

-        vars <- matrixStats::rowVars(logAssay)

-        names(vars) <- rownames(tmpSCE)

-        vars <- sort(vars, decreasing = TRUE)

-        tmpSCE <- tmpSCE[names(vars)[1:nHVG],]

-    }

-    epsilon <- min(nrow(inSCE), epsilon)

-

-    tmpSCE <- zinbwave::zinbwave(tmpSCE, K = nComponents, epsilon = epsilon,

-                                 which_assay = useAssay,

-                                  X = paste('~', batch, sep = ''),

-                                  maxiter.optimize=nIter, verbose = TRUE)

-    reducedDim(inSCE, reducedDimName) <- reducedDim(tmpSCE, 'zinbwave')

-    return(inSCE)

-}

+#' Apply ZINBWaVE Batch effect correction method to SingleCellExperiment object

+#'

+#' A general and flexible zero-inflated negative binomial model that can be

+#' used to provide a low-dimensional representations of scRNAseq data. The

+#' model accounts for zero inflation (dropouts), over-dispersion, and the count

+#' nature of the data. The model also accounts for the difference in library

+#' sizes and optionally for batch effects and/or other covariates.

+#' @param inSCE \linkS4class{SingleCellExperiment} inherited object. Required.

+#' @param useAssay A single character indicating the name of the assay requiring

+#' batch correction. Note that ZINBWaVE works for counts (integer) input rather

+#' than logcounts that other methods prefer. Default \code{"counts"}.

+#' @param batch A single character indicating a field in

+#' \code{\link[SummarizedExperiment]{colData}} that annotates the batches.

+#' Default \code{"batch"}.

+#' @param nHVG An integer. Number of highly variable genes to use when fitting

+#' the model. Default \code{1000L}.

+#' @param nComponents An integer. The number of principle components or

+#' dimensionality to generate in the resulting matrix. Default \code{50L}.

+#' @param nIter An integer, The max number of iterations to perform. Default

+#' \code{10L}.

+#' @param epsilon An integer. Algorithmic parameter. Empirically, a high epsilon

+#' is often required to obtained a good low-level representation. Default

+#' \code{1000L}.

+#' @param reducedDimName A single character. The name for the corrected

+#' low-dimensional representation. Will be saved to \code{reducedDim(inSCE)}.

+#' Default \code{"zinbwave"}.

+#' @return The input \linkS4class{SingleCellExperiment} object with

+#' \code{reducedDim(inSCE, reducedDimName)} updated.

+#' @export

+#' @references Pollen, Alex A et al., 2014

+#' @examples

+#' \dontrun{

+#'     data('sceBatches', package = 'singleCellTK')

+#'     sceCorr <- runZINBWaVE(sceBatches, nIter = 5)

+#' }

+runZINBWaVE <- function(inSCE, useAssay = 'counts', batch = 'batch',

+                        nHVG = 1000L, nComponents = 50L, epsilon = 1000,

+                        nIter = 10L, reducedDimName = 'zinbwave'){

+    ## Input check

+    if(!inherits(inSCE, "SingleCellExperiment")){

+        stop("\"inSCE\" should be a SingleCellExperiment Object.")

+    }

+    if(!useAssay %in% SummarizedExperiment::assayNames(inSCE)) {

+        stop(paste("\"useAssay\" (assay) name: ", useAssay, " not found"))

+    }

+    if(!batch %in% names(SummarizedExperiment::colData(inSCE))){

+        stop(paste("\"batch name:", batch, "not found."))

+    }

+    reducedDimName <- gsub(' ', '_', reducedDimName)

+    nHVG <- as.integer(nHVG)

+    nComponents <- as.integer(nComponents)

+    epsilon <- as.integer(epsilon)

+    nIter <- as.integer(nIter)

+    # Run algorithm

+    ##ZINBWaVE tutorial style of HVG selection

+    if(nHVG < nrow(inSCE)){

+        logAssay <- log1p(SummarizedExperiment::assay(inSCE, useAssay))

+        vars <- matrixStats::rowVars(logAssay)

+        names(vars) <- rownames(inSCE)

+        vars <- sort(vars, decreasing = TRUE)

+        tmpSCE <- inSCE[names(vars)[1:nHVG],]

+    }

+    epsilon <- min(nrow(inSCE), epsilon)

+    print('start!')

+    tmpSCE <- zinbwave::zinbwave(tmpSCE, K = nComponents, epsilon = epsilon,

+                                 which_assay = useAssay,

+                                 X = paste('~', batch, sep = ''),

+                                 maxiter.optimize = nIter, verbose = TRUE)

+    SingleCellExperiment::reducedDim(inSCE, reducedDimName) <-

+        SingleCellExperiment::reducedDim(tmpSCE, 'zinbwave')

+    return(inSCE)

+}

@@ -15,40 +15,39 @@ runLIGER(

 )

 }

 \arguments{

-\item{inSCE}{SingleCellExperiment object. An object that stores your dataset

-and analysis procedures.}

+\item{inSCE}{\linkS4class{SingleCellExperiment} inherited object. Required.}

-\item{useAssay}{character, default `"logcounts"`. A string indicating the name 

-of the assay requiring batch correction in "inSCE", should exist in 

-`assayNames(inSCE)`.}

+\item{useAssay}{A single character indicating the name of the assay requiring

+batch correction. Default \code{"logcounts"}.}

-\item{batch}{character, default `"batch"`. A string indicating the 

-field of `colData(inSCE)` that defines different batches.}

+\item{batch}{A single character indicating a field in

+\code{\link[SummarizedExperiment]{colData}} that annotates the batches.

+Default \code{"batch"}.}

-\item{reducedDimName}{character, default `"LIGER"`. The name for the 

-corrected low-dimensional representation.}

+\item{reducedDimName}{A single character. The name for the corrected

+low-dimensional representation. Will be saved to \code{reducedDim(inSCE)}.

+Default \code{"LIGER"}.}

-\item{nComponents}{integer, default `20L`. Number of principle components or 

-dimensionality (factors, for this algorithm) to generate in the resulting 

-reducedDim.}

+\item{nComponents}{An integer. The number of principle components or

+dimensionality to generate in the resulting matrix. Default \code{20L}.}

-\item{lambda}{numeric, default `5.0`. Algorithmic parameter, the penalty 

-parameter which limits the dataset-specific component of the factorization.}

+\item{lambda}{A numeric scalar. Algorithmic parameter, the penalty

+parameter which limits the dataset-specific component of the factorization.

+Default \code{5.0}.}

-\item{resolution}{numeric, default `1.0`. Algorithmic paramter, the 

-clustering resolution, increasing this increases the number of communities 

-detected.}

+\item{resolution}{A numeric scalar. Algorithmic paramter, the clustering

+resolution, increasing this increases the number of communities detected.

+Default \code{1.0}}

 }

 \value{

-SingleCellExperiment object with `reducedDim(inSCE, reducedDimName)` 

-updated with corrected low-dimentional representation.

+The input \linkS4class{SingleCellExperiment} object with

+\code{reducedDim(inSCE, reducedDimName)} updated.

 }

 \description{

-LIGER relies on integrative non-negative matrix factorization to identify 

+LIGER relies on integrative non-negative matrix factorization to identify

 shared and dataset-specific factors.

 }

 \examples{

- 

 \dontrun{

 data('sceBatches', package = 'singleCellTK')

 sceCorr <- runLIGER(sceBatches)

@@ -7,29 +7,28 @@

 runLimmaBC(inSCE, useAssay = "logcounts", assayName = "LIMMA", batch = "batch")

 }

 \arguments{

-\item{inSCE}{SingleCellExperiment object. An object that stores your dataset

-and analysis procedures.}

+\item{inSCE}{\linkS4class{SingleCellExperiment} inherited object. Required.}

-\item{useAssay}{character, default `"logcounts"`. A string indicating the name 

-of the assay requiring batch correction in "inSCE", should exist in 

-`assayNames(inSCE)`.}

+\item{useAssay}{A single character indicating the name of the assay requiring

+batch correction. Default \code{"logcounts"}.}

-\item{assayName}{character, default `"LIMMA"`. The name for the corrected 

-full-sized expression matrix.}

+\item{assayName}{A single characeter. The name for the corrected assay. Will

+be saved to \code{\link[SummarizedExperiment]{assay}}. Default

+\code{"LIMMA"}.}

-\item{batch}{character, default `"batch"`. A string indicating the 

-field of `colData(inSCE)` that defines different batches.}

+\item{batch}{A single character indicating a field in

+\code{\link[SummarizedExperiment]{colData}} that annotates the batches.

+Default \code{"batch"}.}

 }

 \value{

-SingleCellExperiment object with `assay(inSCE, assayName)` updated 

-with corrected full-sized expression matrix.

+The input \linkS4class{SingleCellExperiment} object with

+\code{assay(inSCE, assayName)} updated.

 }

 \description{

-Limma's batch effect removal function fits a linear model to the data, then 

+Limma's batch effect removal function fits a linear model to the data, then

 removes the component due to the batch effects.

 }

 \examples{

- 

 data('sceBatches', package = 'singleCellTK')

 sceCorr <- runLimmaBC(sceBatches)

 }

@@ -2,7 +2,7 @@

 % Please edit documentation in R/runMNNCorrect.R

 \name{runMNNCorrect}

 \alias{runMNNCorrect}

-\title{Apply the mutual nearest neighbors (MNN) batch effect correction method to 

+\title{Apply the mutual nearest neighbors (MNN) batch effect correction method to

 SingleCellExperiment object}

 \usage{

 runMNNCorrect(

@@ -10,50 +10,50 @@ runMNNCorrect(

   useAssay = "logcounts",

   batch = "batch",

   assayName = "MNN",

-  k = 20,

+  k = 20L,

   sigma = 0.1

 )

 }

 \arguments{

-\item{inSCE}{SingleCellExperiment object. An object that stores your dataset

-and analysis procedures.}

+\item{inSCE}{\linkS4class{SingleCellExperiment} inherited object. Required.}

-\item{useAssay}{character, default `"logcounts"`. A string indicating the name 

-of the assay requiring batch correction in "inSCE", should exist in 

-`assayNames(inSCE)`.}

+\item{useAssay}{A single character indicating the name of the assay requiring

+batch correction. Default \code{"logcounts"}.}

-\item{batch}{character, default `"batch"`. A string indicating the 

-field of `colData(inSCE)` that defines different batches.}

+\item{batch}{A single character indicating a field in

+\code{\link[SummarizedExperiment]{colData}} that annotates the batches.

+Default \code{"batch"}.}

-\item{assayName}{character, default `"MNN"`. The name for the corrected 

-full-sized expression matrix.}

+\item{assayName}{A single characeter. The name for the corrected assay. Will

+be saved to \code{\link[SummarizedExperiment]{assay}}. Default

+\code{"MNN"}.}

-\item{k}{integer, default `20`. Specifies the number of nearest neighbours to 

-consider when defining MNN pairs. This should be interpreted as the minimum 

-frequency of each cell type or state in each batch. Larger values will 

-improve the precision of the correction by increasing the number of MNN 

+\item{k}{An integer. Specifies the number of nearest neighbours to

+consider when defining MNN pairs. This should be interpreted as the minimum

+frequency of each cell type or state in each batch. Larger values will

+improve the precision of the correction by increasing the number of MNN

 pairs, at the cost of reducing accuracy by allowing MNN pairs to form between

-cells of different type.}

+cells of different type. Default \code{20L}.}

-\item{sigma}{Numeric, default `0.1`. Specifies how much information is 

-shared between MNN pairs when computing the batch effect. Larger values will 

-share more information, approaching a global correction for all cells in the 

-same batch. Smaller values allow the correction to vary across cell types, 

-which may be more accurate but comes at the cost of precision.}

+\item{sigma}{A Numeric scalar. Specifies how much information is

+shared between MNN pairs when computing the batch effect. Larger values will

+share more information, approaching a global correction for all cells in the

+same batch. Smaller values allow the correction to vary across cell types,

+which may be more accurate but comes at the cost of precision. Default

+\code{0.1}.}

 }

 \value{

-SingleCellExperiment object with `reducedDim(inSCE, reducedDimName)` 

-updated with corrected low-dimentional representation.

+The input \linkS4class{SingleCellExperiment} object with

+\code{assay(inSCE, assayName)} updated.

 }

 \description{

-MNN is designed for batch correction of single-cell RNA-seq data where the 

-batches are partially confounded with biological conditions of interest. It 

-does so by identifying pairs of MNN in the high-dimensional log-expression 

-space. For each MNN pair, a pairwise correction vector is computed by 

+MNN is designed for batch correction of single-cell RNA-seq data where the

+batches are partially confounded with biological conditions of interest. It

+does so by identifying pairs of MNN in the high-dimensional log-expression

+space. For each MNN pair, a pairwise correction vector is computed by

 applying a Gaussian smoothing kernel with bandwidth `sigma`.

 }

 \examples{

- 

 data('sceBatches', package = 'singleCellTK')

 sceCorr <- runMNNCorrect(sceBatches)

 }

@@ -9,38 +9,38 @@ runSCANORAMA(

   inSCE,

   useAssay = "logcounts",

   batch = "batch",

-  assayName = "SCANORAMA",

   SIGMA = 15,

   ALPHA = 0.1,

-  KNN = 20L

+  KNN = 20L,

+  assayName = "SCANORAMA"

 )

 }

 \arguments{

-\item{inSCE}{SingleCellExperiment object. An object that stores your dataset

-and analysis procedures.}

+\item{inSCE}{\linkS4class{SingleCellExperiment} inherited object. Required.}

-\item{useAssay}{character, default `"logcounts"`. A string indicating the name

-of the assay requiring batch correction in "inSCE", should exist in

-`assayNames(inSCE)`.}

+\item{useAssay}{A single character indicating the name of the assay requiring

+batch correction. Default \code{"logcounts"}.}

-\item{batch}{character, default `"batch"`. A string indicating the

-field of `colData(inSCE)` that defines different batches.}

+\item{batch}{A single character indicating a field in

+\code{\link[SummarizedExperiment]{colData}} that annotates the batches.

+Default \code{"batch"}.}

-\item{assayName}{character, default `"SCANORAMA"`. The name for the

-corrected full-sized expression matrix.}

+\item{SIGMA}{A numeric scalar. Algorithmic parameter, correction smoothing

+parameter on Gaussian kernel. Default \code{15}.}

-\item{SIGMA}{numeric, default `15`. Algorithmic parameter, correction

-smoothing parameter on Gaussian kernel.}

+\item{ALPHA}{A numeric scalar. Algorithmic parameter, alignment score

+minimum cutoff. Default \code{0.1}.}

-\item{ALPHA}{numeric, default `0.1`. Algorithmic parameter, alignment score

-minimum cutoff.}

+\item{KNN}{An integer. Algorithmic parameter, number of nearest neighbors to

+use for matching. Default \code{20L}.}

-\item{KNN}{integer, default `20L`. Algorithmic parameter, number of nearest

-neighbors to use for matching.}

+\item{assayName}{A single characeter. The name for the corrected assay. Will

+be saved to \code{\link[SummarizedExperiment]{assay}}. Default

+\code{"SCANORAMA"}.}

 }

 \value{

-SingleCellExperiment object with `assay(inSCE, assayName)` updated

-with corrected full-sized expression matrix.

+The input \linkS4class{SingleCellExperiment} object with

+\code{assay(inSCE, assayName)} updated.

 }

 \description{

 SCANORAMA is analogous to computer vision algorithms for panorama stitching

@@ -9,40 +9,41 @@ runSCGEN(

   useAssay = "logcounts",

   batch = "batch",

   cellType = "cell_type",

-  assayName = "SCGEN",

-  nEpochs = 50L

+  nEpochs = 50L,

+  assayName = "SCGEN"

 )

 }

 \arguments{

-\item{inSCE}{SingleCellExperiment object. An object that stores your dataset

-and analysis procedures.}

+\item{inSCE}{\linkS4class{SingleCellExperiment} inherited object. Required.}

-\item{useAssay}{character, default `"logcounts"`. A string indicating the name 

-of the assay requiring batch correction in "inSCE", should exist in 

-`assayNames(inSCE)`.}

+\item{useAssay}{A single character indicating the name of the assay requiring

+batch correction. Default \code{"logcounts"}.}

-\item{batch}{character, default `"batch"`. A string indicating the field 

-of `colData(inSCE)` that defines different batches.}

+\item{batch}{A single character indicating a field in

+\code{\link[SummarizedExperiment]{colData}} that annotates the batches.

+Default \code{"batch"}.}

-\item{cellType}{character, default `"cell_type"`. A string indicating the 

-field of `colData(inSCE)` that defines different cell types.}

+\item{cellType}{A single character. A string indicating a field in

+\code{colData(inSCE)} that defines different cell types. Default

+\code{'cell_type'}.}

-\item{assayName}{character, default `"SCGEN"`. The name for the corrected 

-full-sized expression matrix.}

+\item{nEpochs}{An integer. Algorithmic parameter, the number of epochs to

+iterate and optimize network weights. Default \code{50L}.}

-\item{nEpochs}{integer, default `100L`. Algorithmic parameter, number of 

-epochs to iterate and optimize network weights.}

+\item{assayName}{A single characeter. The name for the corrected assay. Will

+be saved to \code{\link[SummarizedExperiment]{assay}}. Default

+\code{"SCGEN"}.}

+}

+\value{

+The input \linkS4class{SingleCellExperiment} object with

+\code{assay(inSCE, assayName)} updated.

 }

 \description{

-scGen is a generative model to predict single-cell perturbation response 

-across cell types, studies and species. It works by combining variational 

+scGen is a generative model to predict single-cell perturbation response

+across cell types, studies and species. It works by combining variational

 autoencoders and latent space vector arithmetics for high-dimensional single-

 cell gene expression data.

 }

-\details{

-Result does not look fine for now. Time consuming also even it allocates 32 

-cores.

-}

 \examples{

 \dontrun{

 data('sceBatches', package = 'singleCellTK')

@@ -12,46 +12,49 @@ runSCMerge(

   seg = NULL,

   kmeansK = NULL,

   cellType = "cell_type",

-  nCores = 1

+  nCores = 1L

 )

 }

 \arguments{

-\item{inSCE}{SingleCellExperiment object. An object that stores your dataset

-and analysis procedures.}

+\item{inSCE}{\linkS4class{SingleCellExperiment} inherited object. Required.}

-\item{useAssay}{character, default `"logcounts"`. A string indicating the name 

-of the assay requiring batch correction in "inSCE", should exist in 

-`assayNames(inSCE)`.}

+\item{useAssay}{A single character indicating the name of the assay requiring

+batch correction. Default \code{"logcounts"}.}

-\item{batch}{character, default `"batch"`. A string indicating the field 

-of `colData(inSCE)` that defines different batches.}

+\item{batch}{A single character indicating a field in

+\code{\link[SummarizedExperiment]{colData}} that annotates the batches.

+Default \code{"batch"}.}

-\item{assayName}{character, default `"scMerge"`. The name for the corrected 

-full-sized expression matrix.}

+\item{assayName}{A single characeter. The name for the corrected assay. Will

+be saved to \code{\link[SummarizedExperiment]{assay}}. Default

+\code{"scMerge"}.}

-\item{seg}{array, default `NULL`. An array of gene names or indices that 

-specifies SEG (Stably Expressed Genes) set as negative control. Pre-defined 

-dataset with human and mouse SEG lists is available to user by running 

-`data('SEG')`.}

+\item{seg}{A vector of gene names or indices that specifies SEG (Stably

+Expressed Genes) set as negative control. Pre-defined dataset with human and

+mouse SEG lists is available to user by running \code{data('SEG')}. Default

+\code{NULL}, and this value will be auto-detected by default with

+\code{\link[scMerge]{scSEGIndex}}.}

-\item{kmeansK}{vector of int, default `NULL`. A vector indicating the 

-kmeans' K-value for each batch, in order to construct pseudo-replicates. The 

-length of `kmeansK` needs to be the same as the number of batches.}

+\item{kmeansK}{An integer vector. Indicating the kmeans' K-value for each

+batch (i.e. how many subclusters in each batch should exist), in order to

+construct pseudo-replicates. The length of code{kmeansK} needs to be the same