R/seurat_integration.R
In connorhknight/IBRAP_no_decontX: Integrated Benchmarking scRNA-seq Analytical Pipeline
Defines functions
perform.seurat.integration
Documented in
perform.seurat.integration
#' @name perform.seurat.cca
#' @aliases perform.seurat.cca
#'
#' @title Performs Seurat Integration
#'
#' @description Performs Seurat integration on the supplied assay names. Results are saved under integration_reductions
#'
#' @param object IBRAP S4 class object
#' @param assay Character. String containing indicating which assay to use
#' @param batch Character. Which column in the metadata defines the batches
#' @param nfeatures Numerical. How many features should be found as integration anchors. Default = 3000
#' @param reduction Character. Which reduction method to use: cca = canonical correlation analysis, rpca = reciprocal PCA, rlsi = Reciprocal LSE. Default = cca
#' @param l2.norm Logical. Perform L2 normalization on the CCA cell embeddings after dimensional reduction. Default = TRUE
#' @param k.anchor Numerical. How many neighbors (k) to use when picking anchors. Default = 5
#' @param k.filter Numerical. How many neighbors (k) to use when filtering anchors. Default = 200
#' @param k.score Numerical. How many neighbors (k) to use when scoring anchors. Default = 30
#' @param max.features. Numerical. The maximum number of features to use when specifying the neighborhood search space in the anchor filtering. Default = 200
#' @param nn.method Character. Method for nearest neighbor finding. Options include: rann, annoy. Default = annoy
#' @param n.tree Numerical. More trees gives higher precision when using annoy approximate nearest neighbor search. Default = 50
#' @param anchor.eps Numerical. Error bound on the neighbor finding algorithm (from RANN/Annoy) when finding integration genes.
#' @param features Character. Vector of features to use when computing the PCA to determine the weights. Only set if you want a different set from those used in the anchor finding process. Default = NULL
#' @param integrate.dims Numerical. Number of dimensions to use in the anchor weighting procedure. Default = 1:30
#' @param k.weight Numerical. Number of neighbors to consider when weighting anchors. Default = 100
#' @param sd.weight Numerical. Controls the bandwidth of the Gaussian kernel for weighting. Default = 1
#' @param sample.tree Character. Specify the order of integration. If NULL, will compute automatically. Default = NULL
#' @param integrate.eps Numerical. Error bound on the neighbor finding algorithm (from RANN)
#' @param save.plot Boolean. Should the automatically genewrated plot be saved? Default = TRUE
#'
#' @return Produces a new 'methods' assay containing normalised, scaled and HVGs.
#'
#' @examples perform.seurat.cca <- function(object = object,
#' assay = c('SCT', 'SCRAN', 'SCANPY'),
#' normalisation.method = c('perform.sct', 'perform.scran', 'perform.scanpy'),
#' batch = original.project,
#' reduction.name.suffix=NULL,
#' nfeatures = 3000,
#' reduction = 'cca',
#' anchors.dims = 1:30,
#' l2.norm = T,
#' k.anchor = 5,
#' k.filter = 200,
#' k.score = 30,
#' max.features = 200,
#' nn.method = 'annoy',
#' n.trees = 50,
#' anchor.eps = 0,
#' features = NULL,
#' integrate.dims = 1:30,
#' k.weight = 100,
#' sd.weight = 1,
#' sample.tree = NULL,
#' integrate.eps = 0)
#'
#' @export
perform.seurat.integration <- function(object,
assay,
normalisation.method,
batch,
reduction.name.suffix=NULL,
nfeatures = 2000,
reduction = 'cca',
anchors.dims = 1:30,
l2.norm = T,
k.anchor = 5,
k.filter = 200,
k.score = 30,
max.features = 200,
nn.method = 'annoy',
n.trees = 50,
anchor.eps = 0,
features = NULL,
integrate.dims = 1:30,
k.weight = 100,
sd.weight = 1,
sample.tree = NULL,
integrate.eps = 0,
save.plot = TRUE) {
if(!is(object = object, class2 = 'IBRAP')) {
stop('object must be of class IBRAP \n')
}
if(!is.character(assay)) {
stop('assay must be character string(s) \n')
} else if(is.character(assay)) {
for(x in assay) {
if(!x %in% names(object@methods)) {
stop(paste0(x, ' is not contained within object@methods \n'))
}
}
}
if(!is.character(normalisation.method)) {
stop('normalisation.method must be character string(s) \n')
} else if(is.character(normalisation.method)) {
for(x in normalisation.method) {
if(!x %in% c('perform.sct', 'perform.scran', 'perform.scanpy', 'perform.tpm')) {
if(isFALSE(x = x)) {
stop(paste0(x, ' function does not exist in the global environment. Does it exist? \n'))
}
}
}
}
if(!is.character(batch)) {
stop('batch must be character string(s) \n')
} else if (is.character(batch)) {
for(x in batch) {
if(!x %in% names(object@sample_metadata)) {
stop(paste0(x, ' is not contained within object@sample_metadata \n'))
}
}
}
if(!is.null(reduction.name.suffix)) {
if(!is.character(reduction.name.suffix)) {
stop('reduction.name.suffix must be a character string \n')
}
}
if(!is.numeric(nfeatures)) {
stop('nfeatures must be cnumerical \n')
}
if(!is.character(reduction)) {
stop('reduction must be character string \n')
} else if(is.character(reduction)) {
if(!reduction %in% c('cca','rpca','rlsi')) {
stop('reduction must be either: cca, rpca or rlsi \n')
}
}
if(!is.logical(l2.norm)) {
stop('l2.norm must be logical. TRUE/FALSE \n')
}
if(!is.numeric(anchors.dims)) {
stop('anchors.dims must be numerical \n')
}
if(!is.numeric(k.anchor)) {
stop('k.anchor must be numerical \n')
}
if(!is.numeric(k.filter)) {
stop('k.filter must be numerical \n')
}
if(!is.numeric(k.score)) {
stop('k.score must be numerical \n')
}
if(!is.numeric(max.features)) {
stop('max.features must be numerical \n')
}
if(!is.character(nn.method)) {
stop('nn.method must be numerical \n')
}
if(!is.numeric(anchor.eps)) {
stop('anchor.eps must be numerical \n')
}
if(!is.null(features)) {
if(!is.character(features)) {
stop('features must be character string(s)')
}
}
if(!is.numeric(integrate.dims)) {
stop('integrate.dims must be numerical \n')
}
if(!is.numeric(k.weight)) {
stop('k.weight must be numerical \n')
}
if(!is.numeric(sd.weight)) {
stop('sd.weight must be numerical \n')
}
if(!is.null(sample.tree)) {
if(!is.character(sample.tree)) {
stop('must be character stirng \n')
}
}
if(!is.numeric(integrate.eps)) {
stop('integrate.eps must be numerical. TRUE/FALSE \n')
}
if(!is.logical(save.plot)) {
stop('save.plot must be boolean. TRUE/FALSE \n')
}
if(length(batch) > 1) {
temp <- function(x) {
return(paste(x, collapse = '_'))
}
df <- object@sample_metadata[,batch]
object2 <- object
object2@sample_metadata$batch <- apply(X = df, MARGIN = 1, FUN = temp)
batch <- 'batch'
} else {
object2 <- object
}
count <- 1
for(a in assay) {
cat(crayon::cyan(paste0(Sys.time(), ': initialising seurat integration for assay: ', a, ' using method: ', reduction, '\n')))
if(normalisation.method[[count]] != 'perform.sct') {
if(normalisation.method[[count]] == 'perform.scran') {
normalise <- function(...) {
perform.scran(...)
}
} else if (normalisation.method[[count]] == 'perform.scanpy') {
normalise <- function(...) {
perform.scanpy(...)
}
} else if (normalisation.method[[count]] == 'perform.tpm') {
normalise <- function(...) {
perform.tpm(...)
}
} else {
stop('normalisation.method must be either: perform.sct, perform.scran, perfrom.scanpy or perform.tpm \n')
}
obj.list <- list()
for(x in unique(object2@sample_metadata[,batch])) {
obj.list[[x]] <- suppressMessages(IBRAP::createIBRAPobject(counts = object@methods$RAW@counts[,object2@sample_metadata[,batch]==x],
original.project = x, add.suffix = F))
}
obj.list <- lapply(X = obj.list, FUN = 'normalise')
new.list <- list()
for(x in names(obj.list)) {
new.list[[x]] <- suppressWarnings(Seurat::CreateSeuratObject(counts = obj.list[[x]]@methods[[2]]@normalised, project = x))
new.list[[x]]@assays$RNA@var.features <- obj.list[[x]]@methods[[2]]@highly.variable.genes
}
features.list <- suppressWarnings(Seurat::SelectIntegrationFeatures(object.list = new.list, nfeatures = nfeatures))
anchors <- suppressWarnings(Seurat::FindIntegrationAnchors(object.list = new.list,
anchor.features = features.list,
reduction = reduction,
scale = T,
l2.norm = l2.norm,
dims = anchors.dims,
k.anchor = k.anchor,
k.filter = k.filter,
k.score = k.score,
max.features = max.features,
nn.method = nn.method,
n.trees = n.trees,
eps = anchor.eps))
to_integrate <- Reduce(intersect, lapply(anchors@object.list, rownames))
combined <- suppressWarnings(Seurat::IntegrateData(anchorset = anchors,
features = features,
features.to.integrate = to_integrate,
dims = integrate.dims,
k.weight = k.weight,
sd.weight = sd.weight,
sample.tree = sample.tree,
preserve.order = T,
eps = integrate.eps))
Seurat::DefaultAssay(combined) <- 'integrated'
combined <- suppressWarnings(Seurat::ScaleData(object = combined))
tmp.obj <- IBRAP::createIBRAPobject(counts = combined@assays$integrated@data,
original.project = 'tmp',
add.suffix = F,
method.name = paste0(a,'_integrated'))
tmp.obj@methods[[paste0(a,'_integrated')]]@norm.scaled <- as.matrix(combined@assays$integrated@scale.data)
tmp.obj@methods[[paste0(a,'_integrated')]]@highly.variable.genes <- combined@assays$integrated@var.features
tmp.obj <- IBRAP::perform.pca(object = tmp.obj, assay = paste0(a,'_integrated'), slot = 'norm.scaled', save.plot = F)
if(is.null(reduction.name.suffix)) {
object@methods[[a]]@integration_reductions[[paste0('seurat', reduction.name.suffix[[count]])]] <- tmp.obj@methods[[1]]@computational_reductions[[1]]
} else if(!is.null(reduction.name.suffix)) {
if(!is.character(reduction.name.suffix)) {
stop('reduction.name.suffix must be character string \n')
} else {
if('_' %in% unlist(strsplit(x = reduction.name.suffix, split = ''))) {
cat(crayon::cyan(paste0(Sys.time(), ': _ cannot be used in reduction.name.suffix, replacing with - \n')))
reduction.name.suffix <- sub(pattern = '_', replacement = '-', x = reduction.name.suffix)
}
object@methods[[a]]@integration_reductions[[paste0('seurat', reduction.name.suffix[[count]])]] <- tmp.obj@methods[[1]]@computational_reductions[[1]]
}
}
} else if(normalisation.method[[count]] == 'perform.sct') {
normalise <- function(...) {
Seurat::SCTransform(...)
}
obj.list <- list()
for(x in unique(object2@sample_metadata[,batch])) {
obj.list[[x]] <- suppressWarnings(Seurat::CreateSeuratObject(counts = object@methods$RAW@counts[,object2@sample_metadata[,batch]==x]))
}
obj.list <- lapply(X = obj.list, FUN = 'normalise')
features <- suppressWarnings(Seurat::SelectIntegrationFeatures(object.list = obj.list, nfeatures = nfeatures))
obj.list <- suppressWarnings(Seurat::PrepSCTIntegration(object.list = obj.list,anchor.features = features))
anchors <- suppressWarnings(Seurat::FindIntegrationAnchors(object.list = obj.list,
normalization.method = "SCT",
anchor.features = features,
reduction = reduction,
scale = F,
l2.norm = l2.norm,
dims = anchors.dims,
k.anchor = k.anchor,
k.filter = k.filter,
k.score = k.score,
max.features = max.features,
nn.method = nn.method,
n.trees = n.trees,
eps = anchor.eps))
to_integrate <- Reduce(intersect, lapply(anchors@object.list, rownames))
combined <- suppressWarnings(Seurat::IntegrateData(anchorset = anchors,
normalization.method = "SCT",
features = features,
features.to.integrate = to_integrate,
dims = integrate.dims,
k.weight = k.weight,
sd.weight = sd.weight,
sample.tree = sample.tree,
preserve.order = T,
eps = integrate.eps))
tmp.obj <- IBRAP::createIBRAPobject(counts = combined@assays$integrated@data,
original.project = 'tmp',
add.suffix = F,
method.name = paste0(a,'_integrated'))
tmp.obj@methods[[paste0(a,'_integrated')]]@norm.scaled <- as.matrix(combined@assays$integrated@scale.data)
tmp.obj@methods[[paste0(a,'_integrated')]]@highly.variable.genes <- combined@assays$integrated@var.features
tmp.obj <- IBRAP::perform.pca(object = tmp.obj, assay = paste0(a,'_integrated'), slot = 'norm.scaled', save.plot = F)
if(is.null(reduction.name.suffix)) {
object@methods[[a]]@integration_reductions[[paste0('seurat', reduction.name.suffix[[count]])]] <- tmp.obj@methods[[1]]@computational_reductions[[1]]
} else if(!is.null(reduction.name.suffix)) {
if(!is.character(reduction.name.suffix)) {
stop('reduction.name.suffix must be character string \n')
} else {
if('_' %in% unlist(strsplit(x = reduction.name.suffix, split = ''))) {
cat(crayon::cyan(paste0(Sys.time(), ': _ cannot be used in reduction.name.suffix, replacing with - \n')))
reduction.name.suffix <- sub(pattern = '_', replacement = '-', x = reduction.name.suffix)
}
object@methods[[a]]@integration_reductions[[paste0('seurat', reduction.name.suffix[[count]])]] <- tmp.obj@methods[[1]]@computational_reductions[[1]]
}
}
}
count <- count + 1
}
return(object)
}
connorhknight/IBRAP_no_decontX documentation built on Feb. 13, 2022, 2:32 p.m.
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
Defines functions perform.seurat.integration
Documented in perform.seurat.integration
#' @name perform.seurat.cca
#' @aliases perform.seurat.cca
#'
#' @title Performs Seurat Integration
#'
#' @description Performs Seurat integration on the supplied assay names. Results are saved under integration_reductions
#'
#' @param object IBRAP S4 class object
#' @param assay Character. String containing indicating which assay to use
#' @param batch Character. Which column in the metadata defines the batches
#' @param nfeatures Numerical. How many features should be found as integration anchors. Default = 3000
#' @param reduction Character. Which reduction method to use: cca = canonical correlation analysis, rpca = reciprocal PCA, rlsi = Reciprocal LSE. Default = cca
#' @param l2.norm Logical. Perform L2 normalization on the CCA cell embeddings after dimensional reduction. Default = TRUE
#' @param k.anchor Numerical. How many neighbors (k) to use when picking anchors. Default = 5
#' @param k.filter Numerical. How many neighbors (k) to use when filtering anchors. Default = 200
#' @param k.score Numerical. How many neighbors (k) to use when scoring anchors. Default = 30
#' @param max.features. Numerical. The maximum number of features to use when specifying the neighborhood search space in the anchor filtering. Default = 200
#' @param nn.method Character. Method for nearest neighbor finding. Options include: rann, annoy. Default = annoy
#' @param n.tree Numerical. More trees gives higher precision when using annoy approximate nearest neighbor search. Default = 50
#' @param anchor.eps Numerical. Error bound on the neighbor finding algorithm (from RANN/Annoy) when finding integration genes.
#' @param features Character. Vector of features to use when computing the PCA to determine the weights. Only set if you want a different set from those used in the anchor finding process. Default = NULL
#' @param integrate.dims Numerical. Number of dimensions to use in the anchor weighting procedure. Default = 1:30
#' @param k.weight Numerical. Number of neighbors to consider when weighting anchors. Default = 100
#' @param sd.weight Numerical. Controls the bandwidth of the Gaussian kernel for weighting. Default = 1
#' @param sample.tree Character. Specify the order of integration. If NULL, will compute automatically. Default = NULL
#' @param integrate.eps Numerical. Error bound on the neighbor finding algorithm (from RANN)
#' @param save.plot Boolean. Should the automatically genewrated plot be saved? Default = TRUE
#'
#' @return Produces a new 'methods' assay containing normalised, scaled and HVGs.
#'
#' @examples perform.seurat.cca <- function(object = object,
#' assay = c('SCT', 'SCRAN', 'SCANPY'),
#' normalisation.method = c('perform.sct', 'perform.scran', 'perform.scanpy'),
#' batch = original.project,
#' reduction.name.suffix=NULL,
#' nfeatures = 3000,
#' reduction = 'cca',
#' anchors.dims = 1:30,
#' l2.norm = T,
#' k.anchor = 5,
#' k.filter = 200,
#' k.score = 30,
#' max.features = 200,
#' nn.method = 'annoy',
#' n.trees = 50,
#' anchor.eps = 0,
#' features = NULL,
#' integrate.dims = 1:30,
#' k.weight = 100,
#' sd.weight = 1,
#' sample.tree = NULL,
#' integrate.eps = 0)
#'
#' @export
perform.seurat.integration <- function(object,
assay,
normalisation.method,
batch,
reduction.name.suffix=NULL,
nfeatures = 2000,
reduction = 'cca',
anchors.dims = 1:30,
l2.norm = T,
k.anchor = 5,
k.filter = 200,
k.score = 30,
max.features = 200,
nn.method = 'annoy',
n.trees = 50,
anchor.eps = 0,
features = NULL,
integrate.dims = 1:30,
k.weight = 100,
sd.weight = 1,
sample.tree = NULL,
integrate.eps = 0,
save.plot = TRUE) {
if(!is(object = object, class2 = 'IBRAP')) {
stop('object must be of class IBRAP \n')
}
if(!is.character(assay)) {
stop('assay must be character string(s) \n')
} else if(is.character(assay)) {
for(x in assay) {
if(!x %in% names(object@methods)) {
stop(paste0(x, ' is not contained within object@methods \n'))
}
}
}
if(!is.character(normalisation.method)) {
stop('normalisation.method must be character string(s) \n')
} else if(is.character(normalisation.method)) {
for(x in normalisation.method) {
if(!x %in% c('perform.sct', 'perform.scran', 'perform.scanpy', 'perform.tpm')) {
if(isFALSE(x = x)) {
stop(paste0(x, ' function does not exist in the global environment. Does it exist? \n'))
}
}
}
}
if(!is.character(batch)) {
stop('batch must be character string(s) \n')
} else if (is.character(batch)) {
for(x in batch) {
if(!x %in% names(object@sample_metadata)) {
stop(paste0(x, ' is not contained within object@sample_metadata \n'))
}
}
}
if(!is.null(reduction.name.suffix)) {
if(!is.character(reduction.name.suffix)) {
stop('reduction.name.suffix must be a character string \n')
}
}
if(!is.numeric(nfeatures)) {
stop('nfeatures must be cnumerical \n')
}
if(!is.character(reduction)) {
stop('reduction must be character string \n')
} else if(is.character(reduction)) {
if(!reduction %in% c('cca','rpca','rlsi')) {
stop('reduction must be either: cca, rpca or rlsi \n')
}
}
if(!is.logical(l2.norm)) {
stop('l2.norm must be logical. TRUE/FALSE \n')
}
if(!is.numeric(anchors.dims)) {
stop('anchors.dims must be numerical \n')
}
if(!is.numeric(k.anchor)) {
stop('k.anchor must be numerical \n')
}
if(!is.numeric(k.filter)) {
stop('k.filter must be numerical \n')
}
if(!is.numeric(k.score)) {
stop('k.score must be numerical \n')
}
if(!is.numeric(max.features)) {
stop('max.features must be numerical \n')
}
if(!is.character(nn.method)) {
stop('nn.method must be numerical \n')
}
if(!is.numeric(anchor.eps)) {
stop('anchor.eps must be numerical \n')
}
if(!is.null(features)) {
if(!is.character(features)) {
stop('features must be character string(s)')
}
}
if(!is.numeric(integrate.dims)) {
stop('integrate.dims must be numerical \n')
}
if(!is.numeric(k.weight)) {
stop('k.weight must be numerical \n')
}
if(!is.numeric(sd.weight)) {
stop('sd.weight must be numerical \n')
}
if(!is.null(sample.tree)) {
if(!is.character(sample.tree)) {
stop('must be character stirng \n')
}
}
if(!is.numeric(integrate.eps)) {
stop('integrate.eps must be numerical. TRUE/FALSE \n')
}
if(!is.logical(save.plot)) {
stop('save.plot must be boolean. TRUE/FALSE \n')
}
if(length(batch) > 1) {
temp <- function(x) {
return(paste(x, collapse = '_'))
}
df <- object@sample_metadata[,batch]
object2 <- object
object2@sample_metadata$batch <- apply(X = df, MARGIN = 1, FUN = temp)
batch <- 'batch'
} else {
object2 <- object
}
count <- 1
for(a in assay) {
cat(crayon::cyan(paste0(Sys.time(), ': initialising seurat integration for assay: ', a, ' using method: ', reduction, '\n')))
if(normalisation.method[[count]] != 'perform.sct') {
if(normalisation.method[[count]] == 'perform.scran') {
normalise <- function(...) {
perform.scran(...)
}
} else if (normalisation.method[[count]] == 'perform.scanpy') {
normalise <- function(...) {
perform.scanpy(...)
}
} else if (normalisation.method[[count]] == 'perform.tpm') {
normalise <- function(...) {
perform.tpm(...)
}
} else {
stop('normalisation.method must be either: perform.sct, perform.scran, perfrom.scanpy or perform.tpm \n')
}
obj.list <- list()
for(x in unique(object2@sample_metadata[,batch])) {
obj.list[[x]] <- suppressMessages(IBRAP::createIBRAPobject(counts = object@methods$RAW@counts[,object2@sample_metadata[,batch]==x],
original.project = x, add.suffix = F))
}
obj.list <- lapply(X = obj.list, FUN = 'normalise')
new.list <- list()
for(x in names(obj.list)) {
new.list[[x]] <- suppressWarnings(Seurat::CreateSeuratObject(counts = obj.list[[x]]@methods[[2]]@normalised, project = x))
new.list[[x]]@assays$RNA@var.features <- obj.list[[x]]@methods[[2]]@highly.variable.genes
}
features.list <- suppressWarnings(Seurat::SelectIntegrationFeatures(object.list = new.list, nfeatures = nfeatures))
anchors <- suppressWarnings(Seurat::FindIntegrationAnchors(object.list = new.list,
anchor.features = features.list,
reduction = reduction,
scale = T,
l2.norm = l2.norm,
dims = anchors.dims,
k.anchor = k.anchor,
k.filter = k.filter,
k.score = k.score,
max.features = max.features,
nn.method = nn.method,
n.trees = n.trees,
eps = anchor.eps))
to_integrate <- Reduce(intersect, lapply(anchors@object.list, rownames))
combined <- suppressWarnings(Seurat::IntegrateData(anchorset = anchors,
features = features,
features.to.integrate = to_integrate,
dims = integrate.dims,
k.weight = k.weight,
sd.weight = sd.weight,
sample.tree = sample.tree,
preserve.order = T,
eps = integrate.eps))
Seurat::DefaultAssay(combined) <- 'integrated'
combined <- suppressWarnings(Seurat::ScaleData(object = combined))
tmp.obj <- IBRAP::createIBRAPobject(counts = combined@assays$integrated@data,
original.project = 'tmp',
add.suffix = F,
method.name = paste0(a,'_integrated'))
tmp.obj@methods[[paste0(a,'_integrated')]]@norm.scaled <- as.matrix(combined@assays$integrated@scale.data)
tmp.obj@methods[[paste0(a,'_integrated')]]@highly.variable.genes <- combined@assays$integrated@var.features
tmp.obj <- IBRAP::perform.pca(object = tmp.obj, assay = paste0(a,'_integrated'), slot = 'norm.scaled', save.plot = F)
if(is.null(reduction.name.suffix)) {
object@methods[[a]]@integration_reductions[[paste0('seurat', reduction.name.suffix[[count]])]] <- tmp.obj@methods[[1]]@computational_reductions[[1]]
} else if(!is.null(reduction.name.suffix)) {
if(!is.character(reduction.name.suffix)) {
stop('reduction.name.suffix must be character string \n')
} else {
if('_' %in% unlist(strsplit(x = reduction.name.suffix, split = ''))) {
cat(crayon::cyan(paste0(Sys.time(), ': _ cannot be used in reduction.name.suffix, replacing with - \n')))
reduction.name.suffix <- sub(pattern = '_', replacement = '-', x = reduction.name.suffix)
}
object@methods[[a]]@integration_reductions[[paste0('seurat', reduction.name.suffix[[count]])]] <- tmp.obj@methods[[1]]@computational_reductions[[1]]
}
}
} else if(normalisation.method[[count]] == 'perform.sct') {
normalise <- function(...) {
Seurat::SCTransform(...)
}
obj.list <- list()
for(x in unique(object2@sample_metadata[,batch])) {
obj.list[[x]] <- suppressWarnings(Seurat::CreateSeuratObject(counts = object@methods$RAW@counts[,object2@sample_metadata[,batch]==x]))
}
obj.list <- lapply(X = obj.list, FUN = 'normalise')
features <- suppressWarnings(Seurat::SelectIntegrationFeatures(object.list = obj.list, nfeatures = nfeatures))
obj.list <- suppressWarnings(Seurat::PrepSCTIntegration(object.list = obj.list,anchor.features = features))
anchors <- suppressWarnings(Seurat::FindIntegrationAnchors(object.list = obj.list,
normalization.method = "SCT",
anchor.features = features,
reduction = reduction,
scale = F,
l2.norm = l2.norm,
dims = anchors.dims,
k.anchor = k.anchor,
k.filter = k.filter,
k.score = k.score,
max.features = max.features,
nn.method = nn.method,
n.trees = n.trees,
eps = anchor.eps))
to_integrate <- Reduce(intersect, lapply(anchors@object.list, rownames))
combined <- suppressWarnings(Seurat::IntegrateData(anchorset = anchors,
normalization.method = "SCT",
features = features,
features.to.integrate = to_integrate,
dims = integrate.dims,
k.weight = k.weight,
sd.weight = sd.weight,
sample.tree = sample.tree,
preserve.order = T,
eps = integrate.eps))
tmp.obj <- IBRAP::createIBRAPobject(counts = combined@assays$integrated@data,
original.project = 'tmp',
add.suffix = F,
method.name = paste0(a,'_integrated'))
tmp.obj@methods[[paste0(a,'_integrated')]]@norm.scaled <- as.matrix(combined@assays$integrated@scale.data)
tmp.obj@methods[[paste0(a,'_integrated')]]@highly.variable.genes <- combined@assays$integrated@var.features
tmp.obj <- IBRAP::perform.pca(object = tmp.obj, assay = paste0(a,'_integrated'), slot = 'norm.scaled', save.plot = F)
if(is.null(reduction.name.suffix)) {
object@methods[[a]]@integration_reductions[[paste0('seurat', reduction.name.suffix[[count]])]] <- tmp.obj@methods[[1]]@computational_reductions[[1]]
} else if(!is.null(reduction.name.suffix)) {
if(!is.character(reduction.name.suffix)) {
stop('reduction.name.suffix must be character string \n')
} else {
if('_' %in% unlist(strsplit(x = reduction.name.suffix, split = ''))) {
cat(crayon::cyan(paste0(Sys.time(), ': _ cannot be used in reduction.name.suffix, replacing with - \n')))
reduction.name.suffix <- sub(pattern = '_', replacement = '-', x = reduction.name.suffix)
}
object@methods[[a]]@integration_reductions[[paste0('seurat', reduction.name.suffix[[count]])]] <- tmp.obj@methods[[1]]@computational_reductions[[1]]
}
}
}
count <- count + 1
}
return(object)
}
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
Embedding an R snippet on your website
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.