| Title: | Scalable differential expression analysis of single cell transcriptomics datasets with complex study designs |
|---|---|
| Description: | Recent advances in single cell/nucleus transcriptomic technology has enabled collection of cohort-scale datasets to study cell type specific gene expression differences associated disease state, stimulus, and genetic regulation. The scale of these data, complex study designs, and low read count per cell mean that characterizing cell type specific molecular mechanisms requires a user-frieldly, purpose-build analytical framework. We have developed the dreamlet package that applies a pseudobulk approach and fits a regression model for each gene and cell cluster to test differential expression across individuals associated with a trait of interest. Use of precision-weighted linear mixed models enables accounting for repeated measures study designs, high dimensional batch effects, and varying sequencing depth or observed cells per biosample. |
| Authors: | Gabriel Hoffman [aut, cre] |
| Maintainer: | Gabriel Hoffman <[email protected]> |
| License: | Artistic-2.0 |
| Version: | 1.5.0 |
| Built: | 2024-12-18 04:16:09 UTC |
| Source: | https://github.com/bioc/dreamlet |
Subset with brackets
Subset with brackets
## S4 method for signature 'dreamletResult,ANY,ANY,ANY' x[i] ## S4 method for signature 'dreamletProcessedData,ANY,ANY,ANY' x[i]## S4 method for signature 'dreamletResult,ANY,ANY,ANY' x[i] ## S4 method for signature 'dreamletProcessedData,ANY,ANY,ANY' x[i]
x |
|
i |
indeces to extract |
entries stored at specified index
Aggregation of single-cell to pseudobulk data for non-count data.
aggregateNonCountSignal( sce, assay = NULL, sample_id = NULL, cluster_id = NULL, min.cells = 10, min.signal = 0.01, min.samples = 4, min.prop = 0.4, verbose = TRUE, BPPARAM = SerialParam(progressbar = verbose) )aggregateNonCountSignal( sce, assay = NULL, sample_id = NULL, cluster_id = NULL, min.cells = 10, min.signal = 0.01, min.samples = 4, min.prop = 0.4, verbose = TRUE, BPPARAM = SerialParam(progressbar = verbose) )
sce |
|
assay |
character string specifying the assay slot to use as
input data. Defaults to the 1st available ( |
sample_id |
character string specifying which variable to use as sample id |
cluster_id |
character string specifying which variable to use as cluster id |
min.cells |
minimum number of observed cells for a sample to be included in the analysis |
min.signal |
minimum signal value for a gene to be considered expressed in a sample. Proper value for this cutoff depends on the type of signal value |
min.samples |
minimum number of samples passing cutoffs for cell cluster to be retained |
min.prop |
minimum proportion of retained samples with non-zero counts for a gene to be |
verbose |
logical. Should information on progress be reported? |
BPPARAM |
a |
The dreamlet workflow can also be applied to non-count data. In this case, a signal is averaged across all cells from a given sample and cell type. Here aggregateNonCountSignal() performs the roles of aggregateToPseudoBulk() followed by processAssays() but using non-count data.
For each cell cluster, samples with at least min.cells are retained. Only clusters with at least min.samples retained samples are kept. Features are retained if they have at least min.signal in at least min.prop fraction of the samples.
The precision of a measurement is the inverse of its sampling variance. The precision weights are computed as 1/sem^2, where sem = sd(signal) / sqrt(n), signal stores the values averaged across cells, and n is the number of cells.
a dreamletProcessedData object
library(muscat) library(SingleCellExperiment) data(example_sce) # create pseudobulk for each sample and cell cluster # using non-count signal pb.signal <- aggregateNonCountSignal(example_sce, assay = "logcounts", cluster_id = "cluster_id", sample_id = "sample_id", verbose = FALSE ) # Differential expression analysis within each assay, # evaluated on the voom normalized data res.dl <- dreamlet(pb.signal, ~group_id)library(muscat) library(SingleCellExperiment) data(example_sce) # create pseudobulk for each sample and cell cluster # using non-count signal pb.signal <- aggregateNonCountSignal(example_sce, assay = "logcounts", cluster_id = "cluster_id", sample_id = "sample_id", verbose = FALSE ) # Differential expression analysis within each assay, # evaluated on the voom normalized data res.dl <- dreamlet(pb.signal, ~group_id)
Aggregation of single-cell to pseudobulk data. Adapted from muscat::aggregateData and has same syntax and results. But can be much faster for SingleCellExperiment backed by H5AD files using on-disk storage.
aggregateToPseudoBulk( x, assay = NULL, sample_id = NULL, cluster_id = NULL, fun = c("sum", "mean", "median", "prop.detected", "num.detected", "sem", "number"), scale = FALSE, verbose = TRUE, BPPARAM = SerialParam(progressbar = verbose), checkValues = TRUE, h5adBlockSizes = 1e+09 )aggregateToPseudoBulk( x, assay = NULL, sample_id = NULL, cluster_id = NULL, fun = c("sum", "mean", "median", "prop.detected", "num.detected", "sem", "number"), scale = FALSE, verbose = TRUE, BPPARAM = SerialParam(progressbar = verbose), checkValues = TRUE, h5adBlockSizes = 1e+09 )
x |
|
assay |
character string specifying the assay slot to use as input data. Defaults to the 1st available ( |
sample_id |
character string specifying which variable to use as sample id |
cluster_id |
character string specifying which variable to use as cluster id |
fun |
a character string.
Specifies the function to use as summary statistic.
Passed to |
scale |
logical. Should pseudo-bulks be scaled with the effective library size & multiplied by 1M? |
verbose |
logical. Should information on progress be reported? |
BPPARAM |
a |
checkValues |
logical. Should we check that signal values are positive integers? |
h5adBlockSizes |
set the automatic block size block size (in bytes) for DelayedArray to read an H5AD file. Larger values use more memory but are faster. |
Adapted from muscat::aggregateData and has similar syntax and same results. This is much faster for SingleCellExperiment backed by H5AD files using DelayedMatrix because this summarizes counts using DelayedMatrixStats. But this function also includes optmizations for sparseMatrix used by Seurat by using sparseMatrixStats.
Keeps variables from colData() that are constant within sample_id. For example, sex will be constant for all cells from the same sample_id, so it is retained as a variable in the pseudobulk result. But number of expressed genes varies across cells within each sample_id, so it is dropped from colData(). Instead the mean value per cell type is stored in metadata(pb)$aggr_means, and these can be included in regression formulas downstream. In that case, the value of the covariates used per sample will depend on the cell type analyzed.
Aggregation parameters (assay, by, fun, scaled) are stored in metadata()$agg_pars, where by = c(cluster_id, sample_id). The number of cells that were aggregated are accessible in int_colData()$n_cells.
Gabriel Hoffman, Helena L Crowell & Mark D Robinson
Crowell, HL, Soneson, C, Germain, P-L, Calini, D, Collin, L, Raposo, C, Malhotra, D & Robinson, MD: Muscat detects subpopulation-specific state transitions from multi-sample multi-condition single-cell transcriptomics data. Nature Communications 11(1):6077 (2020). doi: https://doi.org/10.1038/s41467-020-19894-4
library(muscat) library(SingleCellExperiment) data(example_sce) # create pseudobulk for each sample and cell cluster pb <- aggregateToPseudoBulk(example_sce, assay = "counts", cluster_id = "cluster_id", sample_id = "sample_id", verbose = FALSE ) # pseudobulk data from each cell type # is stored as its own assay pb # aggregate by cluster only, # collapsing all samples into the same pseudobulk pb2 <- aggregateToPseudoBulk(example_sce, cluster_id = "cluster_id", verbose = FALSE) pb2 #library(muscat) library(SingleCellExperiment) data(example_sce) # create pseudobulk for each sample and cell cluster pb <- aggregateToPseudoBulk(example_sce, assay = "counts", cluster_id = "cluster_id", sample_id = "sample_id", verbose = FALSE ) # pseudobulk data from each cell type # is stored as its own assay pb # aggregate by cluster only, # collapsing all samples into the same pseudobulk pb2 <- aggregateToPseudoBulk(example_sce, cluster_id = "cluster_id", verbose = FALSE) pb2 #
Aggregation function for single-cell log-normalized counts to calculate per-sample variance for dreamlet.
aggregateVar( sce, assay = NULL, cluster_id = NULL, sample_id = NULL, min.cells = 10, min.var = 0.01, min.samples = 4, min.prop = 0.4, verbose = TRUE, BPPARAM = SerialParam(progressbar = verbose) )aggregateVar( sce, assay = NULL, cluster_id = NULL, sample_id = NULL, min.cells = 10, min.var = 0.01, min.samples = 4, min.prop = 0.4, verbose = TRUE, BPPARAM = SerialParam(progressbar = verbose) )
sce |
|
assay |
character string specifying the assay slot to use as input data. Defaults to the 1st available ( |
cluster_id |
character string specifying which variable to use as cluster id |
sample_id |
character string specifying which variable to use as sample id |
min.cells |
minimum number of observed cells for a sample to be included in the analysis |
min.var |
minimum variance for a gene to be considered expressed in a sample |
min.samples |
minimum number of samples passing cutoffs for cell cluster to be retained |
min.prop |
minimum proportion of retained samples with non-zero counts for a gene to be |
verbose |
logical. Should information on progress be reported? |
BPPARAM |
a |
The dreamlet workflow can also be applied to model gene expression variance. In this case, a per-sample per-gene variance is calculated across all cells from a given sample and cell type. Here aggregateVar() performs the roles of aggregateToPseudoBulk() followed by processAssays() but using log-normalized count data.
For each cell cluster, samples with at least min.cells are retained. Only clusters with at least min.samples retained samples are kept. Features are retained if they have at least min.var in at least min.prop fraction of the samples.
The precision of a measurement is the inverse of its sampling variance. The precision weights are computed as 1/sem^2, where sem = sd / sqrt(n) and n is the number of cells.
a dreamletProcessedData object
library(muscat) library(SingleCellExperiment) data(example_sce) # Compute variance for each sample and cell cluster pbVar <- aggregateVar(example_sce, assay = "counts", cluster_id = "cluster_id", sample_id = "sample_id", verbose = FALSE )library(muscat) library(SingleCellExperiment) data(example_sce) # Compute variance for each sample and cell cluster pbVar <- aggregateVar(example_sce, assay = "counts", cluster_id = "cluster_id", sample_id = "sample_id", verbose = FALSE )
dreamletResult
Convert list of regression fits to dreamletResult for downstream analysis
as.dreamletResult(fitList, df_details = NULL)as.dreamletResult(fitList, df_details = NULL)
fitList |
list of regression fit with |
df_details |
|
Useful for combining multiple runs of dreamletCompareClusters() into a single dreamletResult for downstream analysis
object of class dreamletResult
library(muscat) library(SingleCellExperiment) data(example_sce) # create pseudobulk for each sample and cell cluster pb <- aggregateToPseudoBulk(example_sce, assay = "counts", cluster_id = "cluster_id", sample_id = "sample_id", verbose = FALSE ) # first comparison ct.pairs <- c("B cells", "CD14+ Monocytes") fit <- dreamletCompareClusters(pb, ct.pairs, method = "fixed") # second comparison ct.pairs2 <- c("B cells", "CD8 T cells") fit2 <- dreamletCompareClusters(pb, ct.pairs2, method = "fixed") # Make a list storing each result with a meaningful name fitList <- list() id <- paste0("[", ct.pairs[1], "]_vs_[", ct.pairs[2], "]") fitList[[id]] <- fit id <- paste0("[", ct.pairs2[1], "]_vs_[", ct.pairs2[2], "]") fitList[[id]] <- fit2 # create a dreamletResult form this list res.compare <- as.dreamletResult(fitList) res.comparelibrary(muscat) library(SingleCellExperiment) data(example_sce) # create pseudobulk for each sample and cell cluster pb <- aggregateToPseudoBulk(example_sce, assay = "counts", cluster_id = "cluster_id", sample_id = "sample_id", verbose = FALSE ) # first comparison ct.pairs <- c("B cells", "CD14+ Monocytes") fit <- dreamletCompareClusters(pb, ct.pairs, method = "fixed") # second comparison ct.pairs2 <- c("B cells", "CD8 T cells") fit2 <- dreamletCompareClusters(pb, ct.pairs2, method = "fixed") # Make a list storing each result with a meaningful name fitList <- list() id <- paste0("[", ct.pairs[1], "]_vs_[", ct.pairs[2], "]") fitList[[id]] <- fit id <- paste0("[", ct.pairs2[1], "]_vs_[", ct.pairs2[2], "]") fitList[[id]] <- fit2 # create a dreamletResult form this list res.compare <- as.dreamletResult(fitList) res.compare
Get assay
Get assay
Get assays by name
## S4 method for signature 'dreamletResult,ANY' assay(x, i, withDimnames = TRUE, ...) ## S4 method for signature 'dreamletProcessedData,ANY' assay(x, i, withDimnames = TRUE, ...) ## S4 method for signature 'vpDF,ANY' assay(x, i, withDimnames = TRUE, ...)## S4 method for signature 'dreamletResult,ANY' assay(x, i, withDimnames = TRUE, ...) ## S4 method for signature 'dreamletProcessedData,ANY' assay(x, i, withDimnames = TRUE, ...) ## S4 method for signature 'vpDF,ANY' assay(x, i, withDimnames = TRUE, ...)
x |
vpDF object |
i |
number indicating index, or string indicating assay |
withDimnames |
not used |
... |
other arguments |
return ith assay
Get assayNames
Get assayNames
Get assayNames
## S4 method for signature 'dreamletResult' assayNames(x, ...) ## S4 method for signature 'dreamletProcessedData' assayNames(x, ...) ## S4 method for signature 'vpDF' assayNames(x, ...)## S4 method for signature 'dreamletResult' assayNames(x, ...) ## S4 method for signature 'dreamletProcessedData' assayNames(x, ...) ## S4 method for signature 'vpDF' assayNames(x, ...)
x |
vpDF object |
... |
additional arguments |
array of assay names
Perform hierarchical clustering on cell types from pseudobulk by aggregating read counts from each cell type.
buildClusterTreeFromPB( pb, method = c("complete", "ward.D", "single", "average", "mcquitty", "median", "centroid", "ward.D2"), dist.method = c("euclidean", "maximum", "manhattan", "canberra", "binary", "minkowski"), assays = assayNames(pb) )buildClusterTreeFromPB( pb, method = c("complete", "ward.D", "single", "average", "mcquitty", "median", "centroid", "ward.D2"), dist.method = c("euclidean", "maximum", "manhattan", "canberra", "binary", "minkowski"), assays = assayNames(pb) )
pb |
|
method |
clustering method for |
dist.method |
distance metric |
assays |
which assays to include |
hierarchical clustering object of class hclust
library(muscat) library(SingleCellExperiment) data(example_sce) # create pseudobulk for each sample and cell cluster pb <- aggregateToPseudoBulk(example_sce, assay = "counts", cluster_id = "cluster_id", sample_id = "sample_id", verbose = FALSE ) # Hierarchical clustering of cell types hcl <- buildClusterTreeFromPB(pb) plot(hcl)library(muscat) library(SingleCellExperiment) data(example_sce) # create pseudobulk for each sample and cell cluster pb <- aggregateToPseudoBulk(example_sce, assay = "counts", cluster_id = "cluster_id", sample_id = "sample_id", verbose = FALSE ) # Hierarchical clustering of cell types hcl <- buildClusterTreeFromPB(pb) plot(hcl)
Extract matrix of cell counts from SingleCellExperiment
cellCounts(x)cellCounts(x)
x |
a |
matrix of cell counts with samples as rows and cell types as columns
computeCellCounts()
library(muscat) library(SingleCellExperiment) data(example_sce) # create pseudobulk for each sample and cell cluster pb <- aggregateToPseudoBulk(example_sce, assay = "counts", cluster_id = "cluster_id", sample_id = "sample_id", verbose = FALSE ) # get matrix of cell counts for each sample cellCounts(pb)library(muscat) library(SingleCellExperiment) data(example_sce) # create pseudobulk for each sample and cell cluster pb <- aggregateToPseudoBulk(example_sce, assay = "counts", cluster_id = "cluster_id", sample_id = "sample_id", verbose = FALSE ) # get matrix of cell counts for each sample cellCounts(pb)
Class cellSpecificityValues cell type specificity values for each gene and cell type
none
For each gene, compute fraction of overall expression attributable to each cell type
cellTypeSpecificity(pb, ...)cellTypeSpecificity(pb, ...)
pb |
|
... |
other arguments passed to |
Sum counts for each cell type, and compute the fraction of counts-per-million attributable to each cell type for each gene
matrix of the fraction of expression attributable to each cell type for each gene.
library(muscat) library(SingleCellExperiment) data(example_sce) # create pseudobulk for each sample and cell cluster pb <- aggregateToPseudoBulk(example_sce, assay = "counts", cluster_id = "cluster_id", sample_id = "sample_id", verbose = FALSE ) # Compute cell type specificity of each gene df <- cellTypeSpecificity(pb) # Violin plot of specificity scores for each cell type # Dashed line indicates genes that are equally expressed # across all cell types. For K cell types, this is 1/K plotViolin(df) # Compute the maximum specificity score for each gene scoreMax <- apply(df, 1, max) head(scoreMax) # For each cell type, get most specific gene genes <- rownames(df)[apply(df, 2, which.max)] # Barplot of 5 genes plotPercentBars(df, genes = genes) # heatmap of 5 genes that are most cell type specific dreamlet::plotHeatmap(df, genes = genes)library(muscat) library(SingleCellExperiment) data(example_sce) # create pseudobulk for each sample and cell cluster pb <- aggregateToPseudoBulk(example_sce, assay = "counts", cluster_id = "cluster_id", sample_id = "sample_id", verbose = FALSE ) # Compute cell type specificity of each gene df <- cellTypeSpecificity(pb) # Violin plot of specificity scores for each cell type # Dashed line indicates genes that are equally expressed # across all cell types. For K cell types, this is 1/K plotViolin(df) # Compute the maximum specificity score for each gene scoreMax <- apply(df, 1, max) head(scoreMax) # For each cell type, get most specific gene genes <- rownames(df)[apply(df, 2, which.max)] # Barplot of 5 genes plotPercentBars(df, genes = genes) # heatmap of 5 genes that are most cell type specific dreamlet::plotHeatmap(df, genes = genes)
Check that variables in formula are present in the data
checkFormula(formula, data)checkFormula(formula, data)
formula |
formula of variables to check |
data |
data.frame storing variables in the formula |
If formula is valid, return TRUE. Else throw error
# Valid formula dreamlet:::checkFormula(~speed, cars) # Not valid formula # dreamlet:::checkFormula( ~ speed + a, cars)# Valid formula dreamlet:::checkFormula(~speed, cars) # Not valid formula # dreamlet:::checkFormula( ~ speed + a, cars)
Get coefficient names
coefNames(obj) ## S4 method for signature 'dreamletResult' coefNames(obj)coefNames(obj) ## S4 method for signature 'dreamletResult' coefNames(obj)
obj |
A |
array storing names of coefficients
library(muscat) library(SingleCellExperiment) data(example_sce) # create pseudobulk for each sample and cell cluster pb <- aggregateToPseudoBulk(example_sce, assay = "counts", cluster_id = "cluster_id", sample_id = "sample_id", verbose = FALSE ) # voom-style normalization res.proc <- processAssays(pb, ~group_id) # Differential expression analysis within each assay, # evaluated on the voom normalized data res.dl <- dreamlet(res.proc, ~group_id) # show coefficients estimated for each cell type coefNames(res.dl)library(muscat) library(SingleCellExperiment) data(example_sce) # create pseudobulk for each sample and cell cluster pb <- aggregateToPseudoBulk(example_sce, assay = "counts", cluster_id = "cluster_id", sample_id = "sample_id", verbose = FALSE ) # voom-style normalization res.proc <- processAssays(pb, ~group_id) # Differential expression analysis within each assay, # evaluated on the voom normalized data res.dl <- dreamlet(res.proc, ~group_id) # show coefficients estimated for each cell type coefNames(res.dl)
dreamletProcessedData
Extract colData from dreamletProcessedData
## S4 method for signature 'dreamletProcessedData' colData(x, ...)## S4 method for signature 'dreamletProcessedData' colData(x, ...)
x |
A |
... |
other arguments |
object from colData field
Set colData of dreamletProcessedData, and check for same dimensions and rownames
## S4 replacement method for signature 'dreamletProcessedData,ANY' colData(x, ...) <- value## S4 replacement method for signature 'dreamletProcessedData,ANY' colData(x, ...) <- value
x |
|
... |
other arguments |
value |
|
none
The posterior probabilities for all genes and conditions is obtained as 1-lFSR. Let prob be an array storing results for one gene. The probability that _no_ conditions in the exclusion set are non-zero is prod(1 - prob[exclude]). The probability that _all_ conditions in the inclusion set are non-zero is prod(prob[include]). The probability that _at least one_ condition in the inclusion set is non-zero is 1 - prod(1 - prob[include]). The composite test is the product of the probabilties computed from the inclusion and exclusion sets.
compositePosteriorTest( x, include, exclude = NULL, test = c("at least 1", "all") )compositePosteriorTest( x, include, exclude = NULL, test = c("at least 1", "all") )
x |
|
include |
array of conditions in the inclusion set |
exclude |
array of conditions in the exclusion set. Defaults to |
test |
evaluate the posterior probability of a non-zero effect in |
Perform composite test evaluating the specificity of an effect. Evalute the posterior probability that an a non-zero effect present in _all_ or _at least one_ condition in the inclusion set, but _no conditions_ in the exclusion set.
run_mash()
library(muscat) library(mashr) library(SingleCellExperiment) data(example_sce) # create pseudobulk for each sample and cell cluster pb <- aggregateToPseudoBulk(example_sce[1:100, ], assay = "counts", cluster_id = "cluster_id", sample_id = "sample_id", verbose = FALSE ) # voom-style normalization res.proc <- processAssays(pb, ~group_id) # Differential expression analysis within each assay, # evaluated on the voom normalized data res.dl <- dreamlet(res.proc, ~group_id) # run MASH model # This can take 10s of minutes on real data # This small datasets should take ~30s res_mash <- run_mash(res.dl, "group_idstim") # Composite test based on posterior probabilities # to identify effect present in *at least 1* monocyte type # and *NO* T-cell type. include <- c("CD14+ Monocytes", "FCGR3A+ Monocytes") exclude <- c("CD4 T cells", "CD8 T cells") # Perform composite test prob <- compositePosteriorTest(res_mash, include, exclude) # examine the lFSR for top gene get_lfsr(res_mash$model)[which.max(prob), , drop = FALSE] # Test if *all* cell types have non-zero effect prob <- compositePosteriorTest(res_mash, assayNames(res.dl))library(muscat) library(mashr) library(SingleCellExperiment) data(example_sce) # create pseudobulk for each sample and cell cluster pb <- aggregateToPseudoBulk(example_sce[1:100, ], assay = "counts", cluster_id = "cluster_id", sample_id = "sample_id", verbose = FALSE ) # voom-style normalization res.proc <- processAssays(pb, ~group_id) # Differential expression analysis within each assay, # evaluated on the voom normalized data res.dl <- dreamlet(res.proc, ~group_id) # run MASH model # This can take 10s of minutes on real data # This small datasets should take ~30s res_mash <- run_mash(res.dl, "group_idstim") # Composite test based on posterior probabilities # to identify effect present in *at least 1* monocyte type # and *NO* T-cell type. include <- c("CD14+ Monocytes", "FCGR3A+ Monocytes") exclude <- c("CD4 T cells", "CD8 T cells") # Perform composite test prob <- compositePosteriorTest(res_mash, include, exclude) # examine the lFSR for top gene get_lfsr(res_mash$model)[which.max(prob), , drop = FALSE] # Test if *all* cell types have non-zero effect prob <- compositePosteriorTest(res_mash, assayNames(res.dl))
Get cell counts with metadata for each sample
computeCellCounts(sce, annotation, sampleIDs)computeCellCounts(sce, annotation, sampleIDs)
sce |
|
annotation |
string indicating column in |
sampleIDs |
string indicating column in |
matrix storing cell counts
library(muscat) library(SingleCellExperiment) data(example_sce) counts <- computeCellCounts(example_sce, "cluster_id", "sample_id") counts[1:4, 1:4]library(muscat) library(SingleCellExperiment) data(example_sce) counts <- computeCellCounts(example_sce, "cluster_id", "sample_id") counts[1:4, 1:4]
Compute normalized counts as log2 counts per million
computeLogCPM( sce, lib.size = colSums2(counts(sce)), prior.count = 2, scaledByLib = FALSE )computeLogCPM( sce, lib.size = colSums2(counts(sce)), prior.count = 2, scaledByLib = FALSE )
sce |
|
lib.size |
library size for each cell |
prior.count |
average count to be added to each observation to avoid taking log of zero |
scaledByLib |
if |
This function gives same result as edgeR::cpm(counts(sce), log=TRUE)
matrix of log CPM values
also edgeR::cpm()
library(muscat) library(SingleCellExperiment) data(example_sce) logcounts(example_sce) <- computeLogCPM(example_sce)library(muscat) library(SingleCellExperiment) data(example_sce) logcounts(example_sce) <- computeLogCPM(example_sce)
Compute normalized counts as counts per million
computeNormCounts(sce)computeNormCounts(sce)
sce |
|
This function gives same result as edgeR::cpm(counts(sce), log=FALSE)
matrix of CPM values
also edgeR::cpm()
library(muscat) library(SingleCellExperiment) data(example_sce) normcounts(example_sce) <- computeNormCounts(example_sce)library(muscat) library(SingleCellExperiment) data(example_sce) normcounts(example_sce) <- computeNormCounts(example_sce)
Extract details from dreamletProcessedData
details(object) ## S4 method for signature 'dreamletProcessedData' details(object) ## S4 method for signature 'dreamletResult' details(object) ## S4 method for signature 'vpDF' details(object)details(object) ## S4 method for signature 'dreamletProcessedData' details(object) ## S4 method for signature 'dreamletResult' details(object) ## S4 method for signature 'vpDF' details(object)
object |
A |
Extract detailed information from some classes
library(muscat) library(SingleCellExperiment) data(example_sce) # create pseudobulk for each sample and cell cluster pb <- aggregateToPseudoBulk(example_sce, assay = "counts", cluster_id = "cluster_id", sample_id = "sample_id", verbose = FALSE ) # voom-style normalization res.proc <- processAssays(pb, ~group_id) # For each cell type, number of samples retained, # and variables retained details(res.proc)library(muscat) library(SingleCellExperiment) data(example_sce) # create pseudobulk for each sample and cell cluster pb <- aggregateToPseudoBulk(example_sce, assay = "counts", cluster_id = "cluster_id", sample_id = "sample_id", verbose = FALSE ) # voom-style normalization res.proc <- processAssays(pb, ~group_id) # For each cell type, number of samples retained, # and variables retained details(res.proc)
Test the association between a covariate of interest and the response's deviation from expectation.
## S4 method for signature 'dreamletResult' diffVar( fit, method = c("AD", "SQ"), scale = c("leverage", "none"), BPPARAM = SerialParam(), ... )## S4 method for signature 'dreamletResult' diffVar( fit, method = c("AD", "SQ"), scale = c("leverage", "none"), BPPARAM = SerialParam(), ... )
fit |
model fit from |
method |
transform the residuals using absolute deviation ("AD") or squared deviation ("SQ"). |
scale |
scale each observation by "leverage", or no scaling ("none") |
BPPARAM |
parameters for parallel evaluation |
... |
other parameters passed to |
This method performs a test of differential variance between two subsets of the data, in a way that generalizes to multiple categories, continuous variables and metrics of spread beyond variance. For the two category test, this method is simular to Levene's test. This model was adapted from Phipson, et al (2014), extended to linear mixed models, and adapted to be compatible with variancePartition::dream() and dreamlet::dreamlet().
This method is composed of multiple steps where 1) a typical linear (mixed) model is fit with dreamlet(), 2) residuals are computed and transformed based on an absolute value or squaring transform, 3) a second regression is performed with dreamlet() to test if a variable is associated with increased deviation from expectation. Both regression take advantage of the dreamlet() linear (mixed) modelling framework followed by empirical Bayes shrinkage that extends the limma::voom() framework.
Note that diffVar() takes the results of the first regression as a parameter to use as a starting point.
Phipson B, Oshlack A (2014). “DiffVar: a new method for detecting differential variability with application to methylation in cancer and aging.” Genome biology, 15(9), 1–16.
variancePartition::diffVar()
variancePartition::diffVar(), missMethyl::diffVar()
library(muscat) library(SingleCellExperiment) data(example_sce) # create pseudobulk for each sample and cell cluster pb <- aggregateToPseudoBulk(example_sce, assay = "counts", cluster_id = "cluster_id", sample_id = "sample_id", verbose = FALSE ) # voom-style normalization res.proc <- processAssays(pb, ~group_id) # Differential expression analysis within each assay, # evaluated on the voom normalized data res.dl <- dreamlet(res.proc, ~group_id) # Differential variance analysis # result is a dreamlet fit res.dvar <- diffVar(res.dl) # Examine results res.dvar # Examine details for each assay details(res.dvar) # show coefficients estimated for each cell type coefNames(res.dvar) # extract results using limma-style syntax # combines all cell types together # adj.P.Val gives study-wide FDR topTable(res.dvar, coef = "group_idstim", number = 3) # Plot top hit to see differential variance # Note that this is a toy example with only 4 samples cellType <- "CD4 T cells" gene <- "DYNLRB1" y <- res.proc[[cellType]]$E[gene, ] x <- colData(res.proc)$group_id boxplot(y ~ x, xlab = "Stimulation status", ylab = "Gene expression", main = paste(cellType, gene) ) #library(muscat) library(SingleCellExperiment) data(example_sce) # create pseudobulk for each sample and cell cluster pb <- aggregateToPseudoBulk(example_sce, assay = "counts", cluster_id = "cluster_id", sample_id = "sample_id", verbose = FALSE ) # voom-style normalization res.proc <- processAssays(pb, ~group_id) # Differential expression analysis within each assay, # evaluated on the voom normalized data res.dl <- dreamlet(res.proc, ~group_id) # Differential variance analysis # result is a dreamlet fit res.dvar <- diffVar(res.dl) # Examine results res.dvar # Examine details for each assay details(res.dvar) # show coefficients estimated for each cell type coefNames(res.dvar) # extract results using limma-style syntax # combines all cell types together # adj.P.Val gives study-wide FDR topTable(res.dvar, coef = "group_idstim", number = 3) # Plot top hit to see differential variance # Note that this is a toy example with only 4 samples cellType <- "CD4 T cells" gene <- "DYNLRB1" y <- res.proc[[cellType]]$E[gene, ] x <- colData(res.proc)$group_id boxplot(y ~ x, xlab = "Stimulation status", ylab = "Gene expression", main = paste(cellType, gene) ) #
Perform differential expression for each assay using linear (mixed) models
dreamlet( x, formula, data = colData(x), assays = assayNames(x), contrasts = NULL, min.cells = 10, robust = FALSE, quiet = FALSE, BPPARAM = SerialParam(), use.eBayes = TRUE, ... ) ## S4 method for signature 'dreamletProcessedData' dreamlet( x, formula, data = colData(x), assays = assayNames(x), contrasts = NULL, min.cells = 10, robust = FALSE, quiet = FALSE, BPPARAM = SerialParam(), use.eBayes = TRUE, ... )dreamlet( x, formula, data = colData(x), assays = assayNames(x), contrasts = NULL, min.cells = 10, robust = FALSE, quiet = FALSE, BPPARAM = SerialParam(), use.eBayes = TRUE, ... ) ## S4 method for signature 'dreamletProcessedData' dreamlet( x, formula, data = colData(x), assays = assayNames(x), contrasts = NULL, min.cells = 10, robust = FALSE, quiet = FALSE, BPPARAM = SerialParam(), use.eBayes = TRUE, ... )
x |
SingleCellExperiment or dreamletProcessedData object |
formula |
regression formula for differential expression analysis |
data |
metadata used in regression formula |
assays |
array of assay names to include in analysis. Defaults to |
contrasts |
character vector specifying contrasts specifying linear combinations of fixed effects to test. This is fed into |
min.cells |
minimum number of observed cells for a sample to be included in the analysis |
robust |
logical, use eBayes method that is robust to outlier genes |
quiet |
show messages |
BPPARAM |
parameters for parallel evaluation |
use.eBayes |
should |
... |
other arguments passed to |
Fit linear (mixed) model on each cell type separately. For advanced use of contrasts see variancePartition::makeContrastsDream() and vignette https://gabrielhoffman.github.io/variancePartition/articles/dream.html#advanced-hypothesis-testing-1.
Object of class dreamletResult storing results for each cell type
variancePartition::dream(), variancePartition::makeContrastsDream()
library(muscat) library(SingleCellExperiment) data(example_sce) # create pseudobulk for each sample and cell cluster pb <- aggregateToPseudoBulk(example_sce, assay = "counts", cluster_id = "cluster_id", sample_id = "sample_id", verbose = FALSE ) # voom-style normalization res.proc <- processAssays(pb, ~group_id) # Differential expression analysis within each assay, # evaluated on the voom normalized data res.dl <- dreamlet(res.proc, ~group_id) # Examine results res.dl # Examine details for each assay details(res.dl) # show coefficients estimated for each cell type coefNames(res.dl) # extract results using limma-style syntax # combines all cell types together # adj.P.Val gives study-wide FDR topTable(res.dl, coef = "group_idstim", number = 3)library(muscat) library(SingleCellExperiment) data(example_sce) # create pseudobulk for each sample and cell cluster pb <- aggregateToPseudoBulk(example_sce, assay = "counts", cluster_id = "cluster_id", sample_id = "sample_id", verbose = FALSE ) # voom-style normalization res.proc <- processAssays(pb, ~group_id) # Differential expression analysis within each assay, # evaluated on the voom normalized data res.dl <- dreamlet(res.proc, ~group_id) # Examine results res.dl # Examine details for each assay details(res.dl) # show coefficients estimated for each cell type coefNames(res.dl) # extract results using limma-style syntax # combines all cell types together # adj.P.Val gives study-wide FDR topTable(res.dl, coef = "group_idstim", number = 3)
Perform differential expression between a pair of assays using linear (mixed) models
dreamletCompareClusters( pb, assays, method = c("fixed", "random", "none"), formula = ~0, collapse = TRUE, min.cells = 10, min.count = 10, min.samples = 4, isCounts = TRUE, normalize.method = "TMM", robust = FALSE, quiet = FALSE, contrasts = c(compare = paste("cellClustertest - cellClusterbaseline")), BPPARAM = SerialParam(), errorsAsWarnings = FALSE, ... )dreamletCompareClusters( pb, assays, method = c("fixed", "random", "none"), formula = ~0, collapse = TRUE, min.cells = 10, min.count = 10, min.samples = 4, isCounts = TRUE, normalize.method = "TMM", robust = FALSE, quiet = FALSE, contrasts = c(compare = paste("cellClustertest - cellClusterbaseline")), BPPARAM = SerialParam(), errorsAsWarnings = FALSE, ... )
pb |
pseudobulk data as SingleCellExperiment object |
assays |
array of two entries specifying assays (i.e. cell clusters) to compare, or a list of two sets of assays. |
method |
account for repeated measures from donors using a "random" effect, a "fixed" effect, or "none" |
formula |
covariates to include in the analysis. |
collapse |
if TRUE (default), combine all cell clusters within the test set, and separately the baseline set. If FALSE, estimate coefficient for each cell cluster and then identify differential expression using linear contrasts with |
min.cells |
minimum number of observed cells for a sample to be included in the analysis |
min.count |
minimum number of reads for a gene to be consider expressed in a sample. Passed to |
min.samples |
minimum number of samples passing cutoffs for cell cluster to be retained |
isCounts |
logical, indicating if data is raw counts |
normalize.method |
normalization method to be used by |
robust |
logical, use eBayes method that is robust to outlier genes |
quiet |
show messages |
contrasts |
cell type is encoded in variable |
BPPARAM |
parameters for parallel evaluation |
errorsAsWarnings |
if |
... |
other arguments passed to |
Analyze pseudobulk data to identify differential gene expression between two cell clusters or sets of clusters while modeling the cross-donor expression variation and other aspects of the study design.
dreamletCompareClusters() is useful for finding genes that are differentially expressed betweeen cell clusters and estimating their fold change. However, the p-values and number of differentially expressed genes are problematic for two reasons, so users must be careful not to overinterpret them:
Cell clusters are typically identified with the same gene expression data used for this differential expression analysis between clusters. The same data is used both for discovery and testing, and this means that the p-values from the differential expression analysis will not be uniform under the null. This will produce a lot of findings with small p-values even in the absence of true biological differences.
The dreamlet package is designed for large datasets with many subjects. The sample sizes from cohort studies are an order of magnitude larger than typical single cell studies. This means that these analyses have huge power to detect even subtle difference in expression between cell clusters. While cluster-specific marker genes are often discovered from an handful of samples, the dreamlet package is applicable to 100s or 1000s of subjects.
method indicates the regression method used to test differential expression between sets of cell clusters. Since the same biosample will usually be represented in both sets of cell clusters, method determines how the paired design is modeled. For method = "mixed", the sample is modeled as a random effect: ~ (1|Sample) + .... For method = "fixed", the sample is modeled as a fixed effect: ~ Sample + .... For method = "none", the pairing is ignored.
When collapse=TRUE (default) combine all cell clusters within the test set, and separately the baseline set, and estimate a coefficient indicating the differential expression between sets for a given gene. If collapse=FALSE, estimate a coefficient for each cell type and then identify differential expression using linear contrasts with variancePartition::makeContrastsDream().
Object of class dreamletResult storing results for each comparison
library(muscat) library(SingleCellExperiment) data(example_sce) # create pseudobulk for each sample and cell cluster pb <- aggregateToPseudoBulk(example_sce, assay = "counts", cluster_id = "cluster_id", sample_id = "sample_id", verbose = FALSE ) # Evaluate the specificity of each gene for each cluster df_cts <- cellTypeSpecificity(pb) # compare first two assays (i.e. cell types) ct.pairs <- c("B cells", "CD14+ Monocytes") # run comparison # use method = 'fixed' here since it is faster fit <- dreamletCompareClusters(pb, ct.pairs, method = "fixed") # Extract top 10 differentially expressed genes # The coefficient 'compare' is the value logFC between test and baseline: # compare = cellClustertest - cellClusterbaseline res <- topTable(fit, coef = "compare", number = 10) # genes with highest logFC are most highly expressed in # B cells compared to CD14+ Monocytes head(res) dreamlet::plotHeatmap(df_cts, genes = rownames(res)[1:5]) # compare B cells versus the rest of the cell types # 'rest' is a keyword indicating all other assays fit <- dreamletCompareClusters(pb, c("B cells", "rest"), method = "fixed") res <- topTable(fit, coef = "compare", number = 10) # genes with highest logFC are most highly expressed in # B cells compared to all others head(res) # Get genes upregulated in B cells idx <- with(res, which(logFC > 0))[1:5] dreamlet::plotHeatmap(df_cts, genes = rownames(res)[idx]) lst <- list( test = c("CD14+ Monocytes", "FCGR3A+ Monocytes"), baseline = c("CD4 T cells", "CD8 T cells") ) # compare 2 monocyte clusters to two T cell clusters fit <- dreamletCompareClusters(pb, lst, method = "fixed") res <- topTable(fit, coef = "compare", number = 10) # genes with highest logFC are most highly expressed in # monocytes compared to T cells head(res) # Get genes upregulated in monocytes idx <- with(res, which(logFC > 0))[1:5] dreamlet::plotHeatmap(df_cts, genes = rownames(res)[idx])library(muscat) library(SingleCellExperiment) data(example_sce) # create pseudobulk for each sample and cell cluster pb <- aggregateToPseudoBulk(example_sce, assay = "counts", cluster_id = "cluster_id", sample_id = "sample_id", verbose = FALSE ) # Evaluate the specificity of each gene for each cluster df_cts <- cellTypeSpecificity(pb) # compare first two assays (i.e. cell types) ct.pairs <- c("B cells", "CD14+ Monocytes") # run comparison # use method = 'fixed' here since it is faster fit <- dreamletCompareClusters(pb, ct.pairs, method = "fixed") # Extract top 10 differentially expressed genes # The coefficient 'compare' is the value logFC between test and baseline: # compare = cellClustertest - cellClusterbaseline res <- topTable(fit, coef = "compare", number = 10) # genes with highest logFC are most highly expressed in # B cells compared to CD14+ Monocytes head(res) dreamlet::plotHeatmap(df_cts, genes = rownames(res)[1:5]) # compare B cells versus the rest of the cell types # 'rest' is a keyword indicating all other assays fit <- dreamletCompareClusters(pb, c("B cells", "rest"), method = "fixed") res <- topTable(fit, coef = "compare", number = 10) # genes with highest logFC are most highly expressed in # B cells compared to all others head(res) # Get genes upregulated in B cells idx <- with(res, which(logFC > 0))[1:5] dreamlet::plotHeatmap(df_cts, genes = rownames(res)[idx]) lst <- list( test = c("CD14+ Monocytes", "FCGR3A+ Monocytes"), baseline = c("CD4 T cells", "CD8 T cells") ) # compare 2 monocyte clusters to two T cell clusters fit <- dreamletCompareClusters(pb, lst, method = "fixed") res <- topTable(fit, coef = "compare", number = 10) # genes with highest logFC are most highly expressed in # monocytes compared to T cells head(res) # Get genes upregulated in monocytes idx <- with(res, which(logFC > 0))[1:5] dreamlet::plotHeatmap(df_cts, genes = rownames(res)[idx])
Class dreamletResult stores results produced by dreamlet() to give a standard interface for downstream analysis
Class dreamletResult stores results produced by dreamlet() to give a standard interface for downstream analysis
none
none
Detect co-linear fixed effects and drop the last one
dropRedundantTerms(formula, data, tol = 0.001)dropRedundantTerms(formula, data, tol = 0.001)
formula |
original formula |
data |
data.frame |
tol |
tolerance to test difference of correlation from 1 or -1 |
a formula, possibly with terms omitted.
# Valid formula dropRedundantTerms(~ group + extra, sleep)# Valid formula dropRedundantTerms(~ group + extra, sleep)
Check if two formulas are equal by evaluating the formulas and extracting terms
equalFormulas(formula1, formula2)equalFormulas(formula1, formula2)
formula1 |
first formula |
formula2 |
second formula |
boolean value indciating of formulas are equivalent
# These formulas are equivalent formula1 <- ~ Size + 1 formula2 <- ~ 1 + Size dreamlet:::equalFormulas(formula1, formula2)# These formulas are equivalent formula1 <- ~ Size + 1 formula2 <- ~ 1 + Size dreamlet:::equalFormulas(formula1, formula2)
colData
Extract normalized expression and colData
Extract normalized (i.e. log2 CPM) expression and colData from dreamletProcessedData
extractData(x, assay, cols = colnames(colData(x)), genes = rownames(x)) ## S4 method for signature 'dreamletProcessedData,character' extractData( x, assay, cols = colnames(colData(x)), genes = rownames(assay(x, assay)) )extractData(x, assay, cols = colnames(colData(x)), genes = rownames(x)) ## S4 method for signature 'dreamletProcessedData,character' extractData( x, assay, cols = colnames(colData(x)), genes = rownames(assay(x, assay)) )
x |
|
assay |
assay to extract |
cols |
columns in |
genes |
genes to extract from |
data.frame or DataFrame of merged expression and colData
library(muscat) library(SingleCellExperiment) data(example_sce) # create pseudobulk for each sample and cell cluster pb <- aggregateToPseudoBulk(example_sce, assay = "counts", cluster_id = "cluster_id", sample_id = "sample_id", verbose = FALSE ) # voom-style normalization res.proc <- processAssays(pb, ~group_id) # Extract all: # Extract tibble of colData merged with expression. # variables and genes are stored as columns, samples as rows df_merge <- extractData(res.proc, "B cells") # first few columns df_merge[, 1:6] # Extract subset: df_merge <- extractData(res.proc, "B cells", cols = "group_id", genes = c("SSU72", "U2AF1")) df_merge # Boxplot of expression boxplot(SSU72 ~ group_id, df_merge) #library(muscat) library(SingleCellExperiment) data(example_sce) # create pseudobulk for each sample and cell cluster pb <- aggregateToPseudoBulk(example_sce, assay = "counts", cluster_id = "cluster_id", sample_id = "sample_id", verbose = FALSE ) # voom-style normalization res.proc <- processAssays(pb, ~group_id) # Extract all: # Extract tibble of colData merged with expression. # variables and genes are stored as columns, samples as rows df_merge <- extractData(res.proc, "B cells") # first few columns df_merge[, 1:6] # Extract subset: df_merge <- extractData(res.proc, "B cells", cols = "group_id", genes = c("SSU72", "U2AF1")) df_merge # Boxplot of expression boxplot(SSU72 ~ group_id, df_merge) #
Perform Variance Partition analysis for each assay
fitVarPart( x, formula, data = colData(x), assays = assayNames(x), quiet = FALSE, BPPARAM = SerialParam(), ... ) ## S4 method for signature 'dreamletProcessedData' fitVarPart( x, formula, data = colData(x), assays = assayNames(x), quiet = FALSE, BPPARAM = SerialParam(), ... )fitVarPart( x, formula, data = colData(x), assays = assayNames(x), quiet = FALSE, BPPARAM = SerialParam(), ... ) ## S4 method for signature 'dreamletProcessedData' fitVarPart( x, formula, data = colData(x), assays = assayNames(x), quiet = FALSE, BPPARAM = SerialParam(), ... )
x |
SingleCellExperiment or dreamletProcessedData object |
formula |
regression formula for differential expression analysis |
data |
metadata used in regression formula |
assays |
array of assay names to include in analysis. Defaults to |
quiet |
show messages |
BPPARAM |
parameters for parallel evaluation |
... |
other arguments passed to |
Object of class vpDF inheriting from DataFrame storing the variance fractions for each gene and cell type.
library(muscat) library(SingleCellExperiment) data(example_sce) # create pseudobulk for each sample and cell cluster pb <- aggregateToPseudoBulk(example_sce, assay = "counts", cluster_id = "cluster_id", sample_id = "sample_id", verbose = FALSE ) # voom-style normalization res.proc <- processAssays(pb, ~group_id) # variance partitioning analysis vp <- fitVarPart(res.proc, ~group_id) # Show variance fractions at the gene-level for each cell type genes <- vp$gene[2:4] plotPercentBars(vp[vp$gene %in% genes, ]) # Summarize variance fractions genome-wide for each cell type plotVarPart(vp)library(muscat) library(SingleCellExperiment) data(example_sce) # create pseudobulk for each sample and cell cluster pb <- aggregateToPseudoBulk(example_sce, assay = "counts", cluster_id = "cluster_id", sample_id = "sample_id", verbose = FALSE ) # voom-style normalization res.proc <- processAssays(pb, ~group_id) # variance partitioning analysis vp <- fitVarPart(res.proc, ~group_id) # Show variance fractions at the gene-level for each cell type genes <- vp$gene[2:4] plotPercentBars(vp[vp$gene %in% genes, ]) # Summarize variance fractions genome-wide for each cell type plotVarPart(vp)
Test if coefficient is different from a specified value
## S4 method for signature 'dreamletResult' getTreat(fit, lfc = log2(1.2), coef = NULL, number = 10, sort.by = "p")## S4 method for signature 'dreamletResult' getTreat(fit, lfc = log2(1.2), coef = NULL, number = 10, sort.by = "p")
fit |
dreamletResult object |
lfc |
a minimum log2-fold-change below which changes not considered scientifically meaningful |
coef |
which coefficient to test |
number |
number of genes to return |
sort.by |
column to sort by |
DataFrame storing hypothesis test for each gene and cell type
limma::topTreat(), variancePartition::getTreat()
library(muscat) library(SingleCellExperiment) data(example_sce) # create pseudobulk for each sample and cell cluster pb <- aggregateToPseudoBulk(example_sce, assay = "counts", cluster_id = "cluster_id", sample_id = "sample_id", verbose = FALSE ) # voom-style normalization res.proc <- processAssays(pb, ~group_id) # Differential expression analysis within each assay, # evaluated on the voom normalized data res.dl <- dreamlet(res.proc, ~group_id) # show coefficients estimated for each cell type coefNames(res.dl) # extract results using limma-style syntax # combines all cell types together # adj.P.Val gives study-wide FDR getTreat(res.dl, coef = "group_idstim", number = 3)library(muscat) library(SingleCellExperiment) data(example_sce) # create pseudobulk for each sample and cell cluster pb <- aggregateToPseudoBulk(example_sce, assay = "counts", cluster_id = "cluster_id", sample_id = "sample_id", verbose = FALSE ) # voom-style normalization res.proc <- processAssays(pb, ~group_id) # Differential expression analysis within each assay, # evaluated on the voom normalized data res.dl <- dreamlet(res.proc, ~group_id) # show coefficients estimated for each cell type coefNames(res.dl) # extract results using limma-style syntax # combines all cell types together # adj.P.Val gives study-wide FDR getTreat(res.dl, coef = "group_idstim", number = 3)
Meta-analysis across multiple studies
meta_analysis( x, method = "FE", group = c("ID", "assay"), control = list(maxiter = 2000) )meta_analysis( x, method = "FE", group = c("ID", "assay"), control = list(maxiter = 2000) )
x |
|
method |
meta-analysis method. Values are fed into |
group |
colums in |
control |
passed to |
'FE': fixed effects meta-analysis
'REML': random effects meta-analysis
'RE2C': joint testing of fixed and random effects
library(dreamlet) library(muscat) data(example_sce) # create pseudobulk for each sample and cell cluster pb <- aggregateToPseudoBulk(example_sce, assay = "counts", cluster_id = "cluster_id", sample_id = "sample_id", verbose = FALSE ) # voom-style normalization # just 'CD14+ Monocytes' for speed res.proc <- processAssays(pb, ~group_id, assays = "CD14+ Monocytes") # dreamlet res.dl <- dreamlet(res.proc, ~group_id) tab1 <- topTable(res.dl, coef = "group_idstim", number = Inf) tab1$Dataset <- "1" # Results from a second cohort # Here, just a copy of the same results for simplicity tab2 <- tab1 tab2$Dataset <- "2" # rbind tab_combined <- rbind(tab1, tab2) # Perform fixed effects meta-analysis res <- meta_analysis(tab_combined, method = "FE") res[1:3, ]library(dreamlet) library(muscat) data(example_sce) # create pseudobulk for each sample and cell cluster pb <- aggregateToPseudoBulk(example_sce, assay = "counts", cluster_id = "cluster_id", sample_id = "sample_id", verbose = FALSE ) # voom-style normalization # just 'CD14+ Monocytes' for speed res.proc <- processAssays(pb, ~group_id, assays = "CD14+ Monocytes") # dreamlet res.dl <- dreamlet(res.proc, ~group_id) tab1 <- topTable(res.dl, coef = "group_idstim", number = Inf) tab1$Dataset <- "1" # Results from a second cohort # Here, just a copy of the same results for simplicity tab2 <- tab1 tab2$Dataset <- "2" # rbind tab_combined <- rbind(tab1, tab2) # Perform fixed effects meta-analysis res <- meta_analysis(tab_combined, method = "FE") res[1:3, ]
dreamletProcessedData
Extract metadata from dreamletProcessedData
## S4 method for signature 'dreamletProcessedData' metadata(x)## S4 method for signature 'dreamletProcessedData' metadata(x)
x |
A dreamletProcessedData object |
object from metadata field
Detect multivariante outliers using Mahalanobis distance using mean and covariance estimated either with standard or robust methods.
outlier(data, robust = FALSE, ...)outlier(data, robust = FALSE, ...)
data |
matrix of data |
robust |
use robust covariance method, defaults to |
... |
arguments passed to |
The distance follow a chisq distrubtion under the null with standard method for mean and covariance. It is approximate if the robust method is used. So use qchisq(p = 0.999 , df = k) to get cutoff to keep 99.9% of samples under the null for data with k=2 columns.
data.frame storing chisq and z-score for each entry indicating deviation from the mean. The z-score is computed by evaluating the p-value of chisq statistic and converting it into a z-score
data <- matrix(rnorm(200), 100, 2) res <- outlier(data) res[1:4,]data <- matrix(rnorm(200), 100, 2) res <- outlier(data) res[1:4,]
Compute outlier score for each sample in each assay using outlier() run on the top principal components. Mahalanobis distance is used for outlier detect and multivariate normal assumption is used to compute p-values
outlierByAssay(object, assays = names(object), nPC = 2, robust = FALSE, ...)outlierByAssay(object, assays = names(object), nPC = 2, robust = FALSE, ...)
object |
|
assays |
assays / cell types to analyze |
nPC |
number of PCs to uses for outlier score with |
robust |
use robust covariance method, defaults to |
... |
arguments passed to |
ID:sample identifier
assay:specify assay
PCs:principal components
chisq:mahalanobis distance that is distributed as chisq(k) k = nPC if the data is multivariate gaussian
z:z-score corresponding to the chisq distance
outlier()
library(muscat) library(SingleCellExperiment) data(example_sce) # create pseudobulk for each sample and cell cluster pb <- aggregateToPseudoBulk(example_sce, assay = "counts", cluster_id = "cluster_id", sample_id = "sample_id", verbose = FALSE ) # voom-style normalization res.proc <- processAssays(pb, ~group_id) # Compute PCs and outlier scores outlierByAssay( res.proc, c("B cells", "CD14+ Monocytes"))library(muscat) library(SingleCellExperiment) data(example_sce) # create pseudobulk for each sample and cell cluster pb <- aggregateToPseudoBulk(example_sce, assay = "counts", cluster_id = "cluster_id", sample_id = "sample_id", verbose = FALSE ) # voom-style normalization res.proc <- processAssays(pb, ~group_id) # Compute PCs and outlier scores outlierByAssay( res.proc, c("B cells", "CD14+ Monocytes"))
Beeswarm plot of effect sizes for each assay, colored by sign and FDR
plotBeeswarm(res.dl, coef, fdr.range = 4, assays = assayNames(res.dl))plotBeeswarm(res.dl, coef, fdr.range = 4, assays = assayNames(res.dl))
res.dl |
|
coef |
coefficient name fed to |
fdr.range |
range for coloring FDR |
assays |
which assays to plot |
ggplot2 of logFC by assay
library(muscat) library(SingleCellExperiment) data(example_sce) # create pseudobulk for each sample and cell cluster pb <- aggregateToPseudoBulk(example_sce, assay = "counts", cluster_id = "cluster_id", sample_id = "sample_id", verbose = FALSE ) # voom-style normalization res.proc <- processAssays(pb, ~group_id) # Differential expression analysis within each assay, # evaluated on the voom normalized data res.dl <- dreamlet(res.proc, ~group_id) # Beeswarm plot of effect sizes for each assay, # colored by sign and FDR plotBeeswarm(res.dl, "group_idstim")library(muscat) library(SingleCellExperiment) data(example_sce) # create pseudobulk for each sample and cell cluster pb <- aggregateToPseudoBulk(example_sce, assay = "counts", cluster_id = "cluster_id", sample_id = "sample_id", verbose = FALSE ) # voom-style normalization res.proc <- processAssays(pb, ~group_id) # Differential expression analysis within each assay, # evaluated on the voom normalized data res.dl <- dreamlet(res.proc, ~group_id) # Beeswarm plot of effect sizes for each assay, # colored by sign and FDR plotBeeswarm(res.dl, "group_idstim")
Bar plot of cell compositions
plotCellComposition(obj, col, width = NULL) ## S4 method for signature 'SingleCellExperiment' plotCellComposition(obj, col, width = NULL) ## S4 method for signature 'matrix' plotCellComposition(obj, col, width = NULL) ## S4 method for signature 'data.frame' plotCellComposition(obj, col, width = NULL)plotCellComposition(obj, col, width = NULL) ## S4 method for signature 'SingleCellExperiment' plotCellComposition(obj, col, width = NULL) ## S4 method for signature 'matrix' plotCellComposition(obj, col, width = NULL) ## S4 method for signature 'data.frame' plotCellComposition(obj, col, width = NULL)
obj |
matrix of [cells] x [samples] or |
col |
array of colors. If missing, use default colors. If |
width |
specify width of bars |
Barplot showing cell fractions
library(muscat) library(SingleCellExperiment) data(example_sce) # create pseudobulk for each sample and cell cluster pb <- aggregateToPseudoBulk(example_sce, assay = "counts", cluster_id = "cluster_id", sample_id = "sample_id", verbose = FALSE ) # show cell composition bar plots plotCellComposition(pb) # extract cell counts df_cellCounts <- cellCounts(pb) # show cell composition bar plots plotCellComposition(df_cellCounts)library(muscat) library(SingleCellExperiment) data(example_sce) # create pseudobulk for each sample and cell cluster pb <- aggregateToPseudoBulk(example_sce, assay = "counts", cluster_id = "cluster_id", sample_id = "sample_id", verbose = FALSE ) # show cell composition bar plots plotCellComposition(pb) # extract cell counts df_cellCounts <- cellCounts(pb) # show cell composition bar plots plotCellComposition(df_cellCounts)
Forest plot
plotForest(x, gene, coef, ...) ## S4 method for signature 'dreamletResult' plotForest(x, gene, coef, assays = names(x), ylim = NULL) ## S4 method for signature 'dreamlet_mash_result' plotForest(x, gene, coef, assays = colnames(x$logFC.original), ylim = NULL)plotForest(x, gene, coef, ...) ## S4 method for signature 'dreamletResult' plotForest(x, gene, coef, assays = names(x), ylim = NULL) ## S4 method for signature 'dreamlet_mash_result' plotForest(x, gene, coef, assays = colnames(x$logFC.original), ylim = NULL)
x |
result from |
gene |
gene to show results for |
coef |
coefficient to test with |
... |
other arguments |
assays |
array of assays to plot |
ylim |
limits for the y axis |
Plot showing effect sizes
library(muscat) library(SingleCellExperiment) data(example_sce) # create pseudobulk for each sample and cell cluster pb <- aggregateToPseudoBulk(example_sce, assay = "counts", cluster_id = "cluster_id", sample_id = "sample_id", verbose = FALSE ) # voom-style normalization res.proc <- processAssays(pb, ~group_id) # Differential expression analysis within each assay, # evaluated on the voom normalized data res.dl <- dreamlet(res.proc, ~group_id) # show coefficients estimated for each cell type coefNames(res.dl) # Show estimated log fold change with in each cell type plotForest(res.dl, gene = "ISG20", coef = "group_idstim")library(muscat) library(SingleCellExperiment) data(example_sce) # create pseudobulk for each sample and cell cluster pb <- aggregateToPseudoBulk(example_sce, assay = "counts", cluster_id = "cluster_id", sample_id = "sample_id", verbose = FALSE ) # voom-style normalization res.proc <- processAssays(pb, ~group_id) # Differential expression analysis within each assay, # evaluated on the voom normalized data res.dl <- dreamlet(res.proc, ~group_id) # show coefficients estimated for each cell type coefNames(res.dl) # Show estimated log fold change with in each cell type plotForest(res.dl, gene = "ISG20", coef = "group_idstim")
Heatmap of genes and assays
plotGeneHeatmap( x, coef, genes, assays = assayNames(x), zmax = NULL, transpose = FALSE ) ## S4 method for signature 'dreamletResult' plotGeneHeatmap( x, coef, genes, assays = assayNames(x), zmax = NULL, transpose = FALSE )plotGeneHeatmap( x, coef, genes, assays = assayNames(x), zmax = NULL, transpose = FALSE ) ## S4 method for signature 'dreamletResult' plotGeneHeatmap( x, coef, genes, assays = assayNames(x), zmax = NULL, transpose = FALSE )
x |
A |
coef |
column number or column name specifying which coefficient or contrast of the linear model is of interest. |
genes |
array of genes to include in plot |
assays |
array of assay names to include in analysis. Defaults to |
zmax |
maximum z.std value |
transpose |
(default: FALSE) Use 'coord_flip()' to flip axies |
Heatmap plot for specified genes and assays
Heatmap plot for specified genes and assays
library(muscat) library(SingleCellExperiment) data(example_sce) # create pseudobulk for each sample and cell cluster pb <- aggregateToPseudoBulk(example_sce, assay = "counts", cluster_id = "cluster_id", sample_id = "sample_id", verbose = FALSE ) # voom-style normalization res.proc <- processAssays(pb, ~group_id) # Differential expression analysis within each assay, # evaluated on the voom normalized data res.dl <- dreamlet(res.proc, ~group_id) # Heatmap for specified subset of genes plotGeneHeatmap(res.dl, coef = "group_idstim", genes = rownames(pb)[1:15])library(muscat) library(SingleCellExperiment) data(example_sce) # create pseudobulk for each sample and cell cluster pb <- aggregateToPseudoBulk(example_sce, assay = "counts", cluster_id = "cluster_id", sample_id = "sample_id", verbose = FALSE ) # voom-style normalization res.proc <- processAssays(pb, ~group_id) # Differential expression analysis within each assay, # evaluated on the voom normalized data res.dl <- dreamlet(res.proc, ~group_id) # Heatmap for specified subset of genes plotGeneHeatmap(res.dl, coef = "group_idstim", genes = rownames(pb)[1:15])
Plot heatmap
plotHeatmap( x, genes = rownames(x), color = "darkblue", assays = colnames(x), useFillScale = TRUE ) ## S4 method for signature 'cellSpecificityValues' plotHeatmap( x, genes = rownames(x), color = "darkblue", assays = colnames(x), useFillScale = TRUE ) ## S4 method for signature 'data.frame' plotHeatmap( x, genes = rownames(x), color = "darkblue", assays = colnames(x), useFillScale = TRUE ) ## S4 method for signature 'matrix' plotHeatmap( x, genes = rownames(x), color = "darkblue", assays = colnames(x), useFillScale = TRUE )plotHeatmap( x, genes = rownames(x), color = "darkblue", assays = colnames(x), useFillScale = TRUE ) ## S4 method for signature 'cellSpecificityValues' plotHeatmap( x, genes = rownames(x), color = "darkblue", assays = colnames(x), useFillScale = TRUE ) ## S4 method for signature 'data.frame' plotHeatmap( x, genes = rownames(x), color = "darkblue", assays = colnames(x), useFillScale = TRUE ) ## S4 method for signature 'matrix' plotHeatmap( x, genes = rownames(x), color = "darkblue", assays = colnames(x), useFillScale = TRUE )
x |
fractions for each gene |
genes |
name of genes to plot |
color |
color of heatmap |
assays |
array of assays to plot |
useFillScale |
default TRUE. add scale_fill_gradient() to plot |
heatmap
library(muscat) library(SingleCellExperiment) data(example_sce) # create pseudobulk for each sample and cell cluster pb <- aggregateToPseudoBulk(example_sce, assay = "counts", cluster_id = "cluster_id", sample_id = "sample_id", verbose = FALSE ) # Compute cell type specificity of each gene df <- cellTypeSpecificity(pb) # For each cell type, get most specific gene genes <- rownames(df)[apply(df, 2, which.max)] # heatmap of 5 genes that are most cell type specific dreamlet::plotHeatmap(df, genes = genes)library(muscat) library(SingleCellExperiment) data(example_sce) # create pseudobulk for each sample and cell cluster pb <- aggregateToPseudoBulk(example_sce, assay = "counts", cluster_id = "cluster_id", sample_id = "sample_id", verbose = FALSE ) # Compute cell type specificity of each gene df <- cellTypeSpecificity(pb) # For each cell type, get most specific gene genes <- rownames(df)[apply(df, 2, which.max)] # heatmap of 5 genes that are most cell type specific dreamlet::plotHeatmap(df, genes = genes)
Compute PCA of gene expression for an assay, and plot samples coloring by outlier score
## S4 method for signature 'list' plotPCA( object, assays = names(object), nPC = 2, robust = FALSE, ..., maxOutlierZ = 20, nrow = 2, size = 2, fdr.cutoff = 0.05 )## S4 method for signature 'list' plotPCA( object, assays = names(object), nPC = 2, robust = FALSE, ..., maxOutlierZ = 20, nrow = 2, size = 2, fdr.cutoff = 0.05 )
object |
|
assays |
assays / cell types to analyze |
nPC |
number of PCs to uses for outlier score with |
robust |
use robust covariance method, defaults to |
... |
arguments passed to |
maxOutlierZ |
cap outlier z-scores at this value for plotting to maintain consistent color scale |
nrow |
number of rows in plot |
size |
size passed to |
fdr.cutoff |
FDR cutoff to determine outlier |
outlierByAssay()
library(muscat) library(SingleCellExperiment) data(example_sce) # create pseudobulk for each sample and cell cluster pb <- aggregateToPseudoBulk(example_sce, assay = "counts", cluster_id = "cluster_id", sample_id = "sample_id", verbose = FALSE ) # voom-style normalization res.proc <- processAssays(pb, ~group_id) # PCA to identify outliers # from normalized expression plotPCA( res.proc, c("B cells", "CD14+ Monocytes")) # Run on regression residuals #----------------------------- # Regression analysis fit = dreamlet(res.proc, ~ group_id) # Extract regression residuals residsObj = residuals(fit) # PCA on residuals plotPCA( residsObj, c("B cells", "CD14+ Monocytes"))library(muscat) library(SingleCellExperiment) data(example_sce) # create pseudobulk for each sample and cell cluster pb <- aggregateToPseudoBulk(example_sce, assay = "counts", cluster_id = "cluster_id", sample_id = "sample_id", verbose = FALSE ) # voom-style normalization res.proc <- processAssays(pb, ~group_id) # PCA to identify outliers # from normalized expression plotPCA( res.proc, c("B cells", "CD14+ Monocytes")) # Run on regression residuals #----------------------------- # Regression analysis fit = dreamlet(res.proc, ~ group_id) # Extract regression residuals residsObj = residuals(fit) # PCA on residuals plotPCA( residsObj, c("B cells", "CD14+ Monocytes"))
Bar plot of variance fractions for a subset of genes
## S4 method for signature 'vpDF' plotPercentBars( x, col = c(ggColorHue(ncol(x) - 3), "grey85"), genes = unique(x$gene), width = NULL, ncol = 3, ... ) ## S4 method for signature 'cellSpecificityValues' plotPercentBars( x, col = ggColorHue(ncol(x)), genes = rownames(x), width = NULL, ... )## S4 method for signature 'vpDF' plotPercentBars( x, col = c(ggColorHue(ncol(x) - 3), "grey85"), genes = unique(x$gene), width = NULL, ncol = 3, ... ) ## S4 method for signature 'cellSpecificityValues' plotPercentBars( x, col = ggColorHue(ncol(x)), genes = rownames(x), width = NULL, ... )
x |
|
col |
color of bars for each variable |
genes |
name of genes to plot |
width |
specify width of bars |
ncol |
number of columns in the plot |
... |
other arguments |
Bar plot showing variance fractions for each gene
library(muscat) library(SingleCellExperiment) data(example_sce) # create pseudobulk for each sample and cell cluster pb <- aggregateToPseudoBulk(example_sce, assay = "counts", cluster_id = "cluster_id", sample_id = "sample_id", verbose = FALSE ) # voom-style normalization res.proc <- processAssays(pb, ~group_id) # variance partitioning analysis vp <- fitVarPart(res.proc, ~group_id) # Show variance fractions at the gene-level for each cell type plotPercentBars(vp, genes = vp$gene[2:4], ncol = 2)library(muscat) library(SingleCellExperiment) data(example_sce) # create pseudobulk for each sample and cell cluster pb <- aggregateToPseudoBulk(example_sce, assay = "counts", cluster_id = "cluster_id", sample_id = "sample_id", verbose = FALSE ) # voom-style normalization res.proc <- processAssays(pb, ~group_id) # variance partitioning analysis vp <- fitVarPart(res.proc, ~group_id) # Show variance fractions at the gene-level for each cell type plotPercentBars(vp, genes = vp$gene[2:4], ncol = 2)
Plot 2D projection (i.e. UMAP, tSNE) for millions of cells efficiently
plotProjection( sce, type, annotation, pointsize = 0, pixels = c(512, 512), legend.position = "none", text = TRUE, order )plotProjection( sce, type, annotation, pointsize = 0, pixels = c(512, 512), legend.position = "none", text = TRUE, order )
sce |
|
type |
field in |
annotation |
column in |
pointsize |
Radius of rasterized point. Use |
pixels |
Vector with X and Y resolution of the raster, default |
legend.position |
legend.position: the position of legends ("none", "left", "right", "bottom", "top", or two-element numeric vector) |
text |
show |
order |
specify order of levels for |
Uses scattermore::geom_scattermore() to plot millions of points efficiently
ggplot2 plot of the projection
library(muscat) library(SingleCellExperiment) data(example_sce) plotProjection(example_sce, "TSNE", "cluster_id", 1)library(muscat) library(SingleCellExperiment) data(example_sce) plotProjection(example_sce, "TSNE", "cluster_id", 1)
Violin plot of variance fraction for each gene and each variable
## S4 method for signature 'DataFrame' plotVarPart( obj, col = c(ggColorHue(base::ncol(obj) - 3), "grey85"), label.angle = 20, main = "", ylab = "", convertToPercent = TRUE, ncol = 3, ... )## S4 method for signature 'DataFrame' plotVarPart( obj, col = c(ggColorHue(base::ncol(obj) - 3), "grey85"), label.angle = 20, main = "", ylab = "", convertToPercent = TRUE, ncol = 3, ... )
obj |
|
col |
vector of colors |
label.angle |
angle of labels on x-axis |
main |
title of plot |
ylab |
text on y-axis |
convertToPercent |
multiply fractions by 100 to convert to percent values |
ncol |
number of columns in the plot |
... |
additional arguments |
Violin plot showing variance fractions
library(muscat) library(SingleCellExperiment) data(example_sce) # create pseudobulk for each sample and cell cluster pb <- aggregateToPseudoBulk(example_sce, assay = "counts", cluster_id = "cluster_id", sample_id = "sample_id", verbose = FALSE ) # voom-style normalization res.proc <- processAssays(pb, ~group_id) # variance partitioning analysis vp <- fitVarPart(res.proc, ~group_id) # Summarize variance fractions genome-wide for each cell type plotVarPart(vp)library(muscat) library(SingleCellExperiment) data(example_sce) # create pseudobulk for each sample and cell cluster pb <- aggregateToPseudoBulk(example_sce, assay = "counts", cluster_id = "cluster_id", sample_id = "sample_id", verbose = FALSE ) # voom-style normalization res.proc <- processAssays(pb, ~group_id) # variance partitioning analysis vp <- fitVarPart(res.proc, ~group_id) # Summarize variance fractions genome-wide for each cell type plotVarPart(vp)
Plot Violins
plotViolin(x, ...) ## S4 method for signature 'cellSpecificityValues' plotViolin(x, assays = colnames(x))plotViolin(x, ...) ## S4 method for signature 'cellSpecificityValues' plotViolin(x, assays = colnames(x))
x |
fractions for each gene |
... |
other arguments |
assays |
array of assays to plot |
Violin plot
library(muscat) library(SingleCellExperiment) data(example_sce) # create pseudobulk for each sample and cell cluster pb <- aggregateToPseudoBulk(example_sce, assay = "counts", cluster_id = "cluster_id", sample_id = "sample_id", verbose = FALSE ) # Compute cell type specificity of each gene df <- cellTypeSpecificity(pb) # Violin plot of specificity scores for each cell type # Dashed line indicates genes that are equally expressed # across all cell types. For K cell types, this is 1/K plotViolin(df)library(muscat) library(SingleCellExperiment) data(example_sce) # create pseudobulk for each sample and cell cluster pb <- aggregateToPseudoBulk(example_sce, assay = "counts", cluster_id = "cluster_id", sample_id = "sample_id", verbose = FALSE ) # Compute cell type specificity of each gene df <- cellTypeSpecificity(pb) # Violin plot of specificity scores for each cell type # Dashed line indicates genes that are equally expressed # across all cell types. For K cell types, this is 1/K plotViolin(df)
Volcano plot for each cell type
plotVolcano( x, coef, nGenes = 5, size = 12, minp = 9.99999999999997e-311, cutoff = 0.05, ncol = 3, ... ) ## S4 method for signature 'list' plotVolcano( x, coef, nGenes = 5, size = 12, minp = 9.99999999999997e-311, cutoff = 0.05, ncol = 3, assays = names(x), ... ) ## S4 method for signature 'MArrayLM' plotVolcano( x, coef, nGenes = 5, size = 12, minp = 9.99999999999997e-311, cutoff = 0.05, ncol = 3, ... ) ## S4 method for signature 'dreamlet_mash_result' plotVolcano( x, coef, nGenes = 5, size = 12, minp = 1e-16, cutoff = 0.05, ncol = 3, assays = colnames(x$logFC.original), ... )plotVolcano( x, coef, nGenes = 5, size = 12, minp = 9.99999999999997e-311, cutoff = 0.05, ncol = 3, ... ) ## S4 method for signature 'list' plotVolcano( x, coef, nGenes = 5, size = 12, minp = 9.99999999999997e-311, cutoff = 0.05, ncol = 3, assays = names(x), ... ) ## S4 method for signature 'MArrayLM' plotVolcano( x, coef, nGenes = 5, size = 12, minp = 9.99999999999997e-311, cutoff = 0.05, ncol = 3, ... ) ## S4 method for signature 'dreamlet_mash_result' plotVolcano( x, coef, nGenes = 5, size = 12, minp = 1e-16, cutoff = 0.05, ncol = 3, assays = colnames(x$logFC.original), ... )
x |
result from |
coef |
coefficient to test with |
nGenes |
number of genes to highlight in each volcano plot |
size |
text size |
minp |
minimum p-value to show on the y-axis |
cutoff |
adj.P.Val cutoff to distinguish significant from non-significant genes |
ncol |
number of columns in the plot |
... |
arguments passed to |
assays |
which assays to plot |
Volcano plot for each cell type
library(muscat) library(SingleCellExperiment) data(example_sce) # create pseudobulk for each sample and cell cluster pb <- aggregateToPseudoBulk(example_sce, assay = "counts", cluster_id = "cluster_id", sample_id = "sample_id", verbose = FALSE ) # voom-style normalization res.proc <- processAssays(pb, ~group_id) # Differential expression analysis within each assay, # evaluated on the voom normalized data res.dl <- dreamlet(res.proc, ~group_id) # show coefficients estimated for each cell type coefNames(res.dl) # volcano plot for each cell type plotVolcano(res.dl, coef = "group_idstim") # volcano plot for first two cell types plotVolcano(res.dl[1:2], coef = "group_idstim")library(muscat) library(SingleCellExperiment) data(example_sce) # create pseudobulk for each sample and cell cluster pb <- aggregateToPseudoBulk(example_sce, assay = "counts", cluster_id = "cluster_id", sample_id = "sample_id", verbose = FALSE ) # voom-style normalization res.proc <- processAssays(pb, ~group_id) # Differential expression analysis within each assay, # evaluated on the voom normalized data res.dl <- dreamlet(res.proc, ~group_id) # show coefficients estimated for each cell type coefNames(res.dl) # volcano plot for each cell type plotVolcano(res.dl, coef = "group_idstim") # volcano plot for first two cell types plotVolcano(res.dl[1:2], coef = "group_idstim")
Plot voom curves from each cell type
plotVoom(x, ncol = 3, alpha = 0.5, ...) ## S4 method for signature 'dreamletProcessedData' plotVoom(x, ncol = 3, alpha = 0.5, assays = names(x)) ## S4 method for signature 'EList' plotVoom(x, ncol = 3, alpha = 0.5)plotVoom(x, ncol = 3, alpha = 0.5, ...) ## S4 method for signature 'dreamletProcessedData' plotVoom(x, ncol = 3, alpha = 0.5, assays = names(x)) ## S4 method for signature 'EList' plotVoom(x, ncol = 3, alpha = 0.5)
x |
dreamletProcessedData |
ncol |
number of columns in the plot |
alpha |
transparency of points |
... |
other arguments |
assays |
which assays to plot |
Plot of mean-variance trend
library(muscat) library(SingleCellExperiment) data(example_sce) # create pseudobulk for each sample and cell cluster pb <- aggregateToPseudoBulk(example_sce, assay = "counts", cluster_id = "cluster_id", sample_id = "sample_id", verbose = FALSE ) # voom-style normalization res.proc <- processAssays(pb, ~group_id) # Show mean-variance trend from voom plotVoom(res.proc) # plot for first two cell types plotVoom(res.proc[1:2])library(muscat) library(SingleCellExperiment) data(example_sce) # create pseudobulk for each sample and cell cluster pb <- aggregateToPseudoBulk(example_sce, assay = "counts", cluster_id = "cluster_id", sample_id = "sample_id", verbose = FALSE ) # voom-style normalization res.proc <- processAssays(pb, ~group_id) # Show mean-variance trend from voom plotVoom(res.proc) # plot for first two cell types plotVoom(res.proc[1:2])
Print object
Print object
## S4 method for signature 'dreamletResult' print(x, ...) ## S4 method for signature 'dreamletProcessedData' print(x, ...)## S4 method for signature 'dreamletResult' print(x, ...) ## S4 method for signature 'dreamletProcessedData' print(x, ...)
x |
|
... |
other arguments |
print data stored in object
For raw counts, estimate precision weights using linear mixed model weighting by number of cells observed for each sample. For normalized data, only weight by number of cells.
processAssays( sceObj, formula, assays = assayNames(sceObj), min.cells = 5, min.count = 5, min.samples = 4, min.prop = 0.4, isCounts = TRUE, normalize.method = "TMM", span = "auto", quiet = FALSE, weightsList = NULL, BPPARAM = SerialParam(), ... )processAssays( sceObj, formula, assays = assayNames(sceObj), min.cells = 5, min.count = 5, min.samples = 4, min.prop = 0.4, isCounts = TRUE, normalize.method = "TMM", span = "auto", quiet = FALSE, weightsList = NULL, BPPARAM = SerialParam(), ... )
sceObj |
SingleCellExperiment object |
formula |
regression formula for differential expression analysis |
assays |
array of assay names to include in analysis. Defaults to |
min.cells |
minimum number of observed cells for a sample to be included in the analysis |
min.count |
minimum number of reads for a gene to be considered expressed in a sample. Passed to |
min.samples |
minimum number of samples passing cutoffs for cell cluster to be retained |
min.prop |
minimum proportion of retained samples with non-zero counts for a gene to be retained |
isCounts |
logical, indicating if data is raw counts |
normalize.method |
normalization method to be used by |
span |
Lowess smoothing parameter using by |
quiet |
show messages |
weightsList |
list storing matrix of precision weights for each cell type. If |
BPPARAM |
parameters for parallel evaluation |
... |
other arguments passed to |
For each cell cluster, samples with at least min.cells are retained. Only clusters with at least min.samples retained samples are kept. Genes are retained if they have at least min.count reads in at least min.prop fraction of the samples. Current values are reasonable defaults, since genes that don't pass these cutoffs are very underpowered for differential expression analysis and only increase the multiple testing burden. But values of min.cells = 2 and min.count = 2 are also reasonable to include more genes in the analysis.
The precision weights are estimated using the residuals fit from the specified formula. These weights are robust to changes in the formula as long as the major variables explaining the highest fraction of the variance are included.
If weightsList is NULL, precision weights are set to 1 internally.
Object of class dreamletProcessedData storing voom-style normalized expression data
voomWithDreamWeights()
library(muscat) library(SingleCellExperiment) data(example_sce) # create pseudobulk for each sample and cell cluster pb <- aggregateToPseudoBulk(example_sce, assay = "counts", cluster_id = "cluster_id", sample_id = "sample_id", verbose = FALSE ) # voom-style normalization res.proc <- processAssays(pb, ~group_id) # Differential expression analysis within each assay, # evaluated on the voom normalized data res.dl <- dreamlet(res.proc, ~group_id) #library(muscat) library(SingleCellExperiment) data(example_sce) # create pseudobulk for each sample and cell cluster pb <- aggregateToPseudoBulk(example_sce, assay = "counts", cluster_id = "cluster_id", sample_id = "sample_id", verbose = FALSE ) # voom-style normalization res.proc <- processAssays(pb, ~group_id) # Differential expression analysis within each assay, # evaluated on the voom normalized data res.dl <- dreamlet(res.proc, ~group_id) #
For raw counts, filter genes and samples, then estimate precision weights using linear mixed model weighting by number of cells observed for each sample. For normalized data, only weight by number of cells
processOneAssay( y, formula, data, n.cells, min.cells = 5, min.count = 2, min.samples = 4, min.prop = 0.4, min.total.count = 15, isCounts = TRUE, normalize.method = "TMM", span = "auto", quiet = TRUE, weights = NULL, rescaleWeightsAfter = FALSE, BPPARAM = SerialParam(), ... )processOneAssay( y, formula, data, n.cells, min.cells = 5, min.count = 2, min.samples = 4, min.prop = 0.4, min.total.count = 15, isCounts = TRUE, normalize.method = "TMM", span = "auto", quiet = TRUE, weights = NULL, rescaleWeightsAfter = FALSE, BPPARAM = SerialParam(), ... )
y |
matrix of counts or log2 CPM |
formula |
regression formula for differential expression analysis |
data |
metadata used in regression formula |
n.cells |
array of cell count for each sample |
min.cells |
minimum number of observed cells for a sample to be included in the analysis |
min.count |
minimum number of reads for a gene to be considered expressed in a sample. Passed to |
min.samples |
minimum number of samples passing cutoffs for cell cluster to be retained |
min.prop |
minimum proportion of retained samples with non-zero counts |
min.total.count |
minimum total count required per gene for inclusion |
isCounts |
logical, indicating if data is raw counts |
normalize.method |
normalization method to be used by |
span |
Lowess smoothing parameter using by |
quiet |
show messages |
weights |
matrix of precision weights |
rescaleWeightsAfter |
default = FALSE, should the output weights be scaled by the input weights |
BPPARAM |
parameters for parallel evaluation |
... |
other arguments passed to |
EList object storing log2 CPM and precision weights
processAssays()
Remove constant terms from formula. Also remove categorical variables with a max of one example per category
removeConstantTerms(formula, data)removeConstantTerms(formula, data)
formula |
original formula |
data |
data.frame |
Adapted from MoEClust::drop_constants
a formula, possibly with terms omitted.
# Valid formula removeConstantTerms(~ group + extra, sleep) # there is no variation in 'group' in this dataset removeConstantTerms(~ group + extra, sleep[1:3, ])# Valid formula removeConstantTerms(~ group + extra, sleep) # there is no variation in 'group' in this dataset removeConstantTerms(~ group + extra, sleep[1:3, ])
dreamletResult
Extract residuals from dreamletResult
## S4 method for signature 'dreamletResult' residuals(object, y, ..., type = c("response", "pearson"))## S4 method for signature 'dreamletResult' residuals(object, y, ..., type = c("response", "pearson"))
object |
|
y |
|
... |
other arguments |
type |
compute either |
"response" residuals are the typical residuals returned from lm(). "pearson" residuals divides each residual value by its estimated standard error. This requires specifying y
residuals from model fit
library(muscat) library(SingleCellExperiment) data(example_sce) # create pseudobulk for each sample and cell cluster pb <- aggregateToPseudoBulk(example_sce, assay = "counts", cluster_id = "cluster_id", sample_id = "sample_id", verbose = FALSE ) # voom-style normalization res.proc <- processAssays(pb, ~group_id) # Differential expression analysis within each assay, # evaluated on the voom normalized data res.dl <- dreamlet(res.proc, ~group_id) # extract typical residuals for each assay (i.e. cell type) # Return list with entry for each assay with for retained samples and genes resid.lst <- residuals(res.dl) # Get Pearson residuals: # typical residuals scaled by the standard deviation residPearson.lst <- residuals(res.dl, res.proc, type = "pearson")library(muscat) library(SingleCellExperiment) data(example_sce) # create pseudobulk for each sample and cell cluster pb <- aggregateToPseudoBulk(example_sce, assay = "counts", cluster_id = "cluster_id", sample_id = "sample_id", verbose = FALSE ) # voom-style normalization res.proc <- processAssays(pb, ~group_id) # Differential expression analysis within each assay, # evaluated on the voom normalized data res.dl <- dreamlet(res.proc, ~group_id) # extract typical residuals for each assay (i.e. cell type) # Return list with entry for each assay with for retained samples and genes resid.lst <- residuals(res.dl) # Get Pearson residuals: # typical residuals scaled by the standard deviation residPearson.lst <- residuals(res.dl, res.proc, type = "pearson")
Run mash analysis on dreamlet results
run_mash(fit, coefList)run_mash(fit, coefList)
fit |
result from |
coefList |
coefficient to be analyzed. Assumes 1) the null distribution of the two coefficients is simular, 2) the effects sizes are on the same scale, and 3) the effect estimates should be shrunk towards each other. If these are not satisfied, run separately on each coefficient |
Apply mashr analysis (Urbut et al. 2019) on the joint set of coefficients for each gene and cell type. mashr is a Bayesian statistical method that borrows strength across tests (i.e. genes and cell types) by learning the distribution of non-zero effects based the obesrved logFC and standard errors. The method then estimates the posterior distributions of each coefficient based on the observed value and the genome-wide emprical distribution.
mashr has been previously applied to differential expression in GTEx data using multiple tissues from the same set of donors (Oliva et al. 2020).
In single cell data, a given gene is often not sufficiently expressed in all cell types. So it is not evaluated in a subsets of cell types, and its coefficient value is NA. Since mashr assumes coefficients and standard errors for every gene and cell type pair, entries with these missing values are set to have coef = 0, and se = 1e6. The output of mashr is then modified to set the corresponding values to NA, to avoid nonsensical results downstream.
a list storing the mashr model as model and the original coefficients as logFC.original
Oliva M, Munoz-Aguirre M, Kim-Hellmuth S, Wucher V, Gewirtz AD, Cotter DJ, Parsana P, Kasela S, Balliu B, Vinuela A, others (2020).
“The impact of sex on gene expression across human tissues.”
Science, 369(6509), eaba3066.
https://doi.org/10.1126/science.aba3066.
Urbut SM, Wang G, Carbonetto P, Stephens M (2019).
“Flexible statistical methods for estimating and testing effects in genomic studies with multiple conditions.”
Nature genetics, 51(1), 187–195.
https://doi.org/10.1038/s41588-018-0268-8.
mashr::mash_estimate_corr_em(), mashr::cov_canonical, mashr::mash_set_data
library(muscat) library(mashr) library(SingleCellExperiment) data(example_sce) # create pseudobulk for each sample and cell cluster pb <- aggregateToPseudoBulk(example_sce[1:100, ], assay = "counts", cluster_id = "cluster_id", sample_id = "sample_id", verbose = FALSE ) # voom-style normalization res.proc <- processAssays(pb, ~group_id) # Differential expression analysis within each assay, # evaluated on the voom normalized data res.dl <- dreamlet(res.proc, ~group_id) # run MASH model # This can take 10s of minutes on real data # This small datasets should take ~30s res_mash <- run_mash(res.dl, "group_idstim") # extract statistics from mashr model # NA values indicate genes not sufficiently expressed # in a given cell type # original logFC head(res_mash$logFC.original) # posterior mean for logFC head(get_pm(res_mash$model)) # how many gene-by-celltype tests are significant # i.e. if a gene is significant in 2 celltypes, it is counted twice table(get_lfsr(res_mash$model) < 0.05, useNA = "ifany") # how many genes are significant in at least one cell type table(apply(get_lfsr(res_mash$model), 1, min, na.rm = TRUE) < 0.05) # how many genes are significant in each cell type apply(get_lfsr(res_mash$model), 2, function(x) sum(x < 0.05, na.rm = TRUE)) # examine top set of genes # which genes are significant in at least 1 cell type sort(names(get_significant_results(res_mash$model)))[1:10] # Lets examine ENO1 # There is a lot of variation in the raw logFC res_mash$logFC.original["ENO1", ] # posterior mean after borrowing across cell type and genes get_pm(res_mash$model)["ENO1", ] # forest plot based on mashr results plotForest(res_mash, "ENO1") # volcano plot based on mashr results # yaxis uses local false sign rate (lfsr) plotVolcano(res_mash) # Comment out to reduce package runtime # gene set analysis using mashr results # library(zenith) # go.gs = get_GeneOntology("CC", to="SYMBOL") # df_gs = zenith_gsa(res_mash, go.gs) # Heatmap of results # plotZenithResults(df_gs, 2, 1)library(muscat) library(mashr) library(SingleCellExperiment) data(example_sce) # create pseudobulk for each sample and cell cluster pb <- aggregateToPseudoBulk(example_sce[1:100, ], assay = "counts", cluster_id = "cluster_id", sample_id = "sample_id", verbose = FALSE ) # voom-style normalization res.proc <- processAssays(pb, ~group_id) # Differential expression analysis within each assay, # evaluated on the voom normalized data res.dl <- dreamlet(res.proc, ~group_id) # run MASH model # This can take 10s of minutes on real data # This small datasets should take ~30s res_mash <- run_mash(res.dl, "group_idstim") # extract statistics from mashr model # NA values indicate genes not sufficiently expressed # in a given cell type # original logFC head(res_mash$logFC.original) # posterior mean for logFC head(get_pm(res_mash$model)) # how many gene-by-celltype tests are significant # i.e. if a gene is significant in 2 celltypes, it is counted twice table(get_lfsr(res_mash$model) < 0.05, useNA = "ifany") # how many genes are significant in at least one cell type table(apply(get_lfsr(res_mash$model), 1, min, na.rm = TRUE) < 0.05) # how many genes are significant in each cell type apply(get_lfsr(res_mash$model), 2, function(x) sum(x < 0.05, na.rm = TRUE)) # examine top set of genes # which genes are significant in at least 1 cell type sort(names(get_significant_results(res_mash$model)))[1:10] # Lets examine ENO1 # There is a lot of variation in the raw logFC res_mash$logFC.original["ENO1", ] # posterior mean after borrowing across cell type and genes get_pm(res_mash$model)["ENO1", ] # forest plot based on mashr results plotForest(res_mash, "ENO1") # volcano plot based on mashr results # yaxis uses local false sign rate (lfsr) plotVolcano(res_mash) # Comment out to reduce package runtime # gene set analysis using mashr results # library(zenith) # go.gs = get_GeneOntology("CC", to="SYMBOL") # df_gs = zenith_gsa(res_mash, go.gs) # Heatmap of results # plotZenithResults(df_gs, 2, 1)
Get error text
seeErrors(obj) ## S4 method for signature 'dreamletResult' seeErrors(obj) ## S4 method for signature 'dreamletProcessedData' seeErrors(obj) ## S4 method for signature 'vpDF' seeErrors(obj)seeErrors(obj) ## S4 method for signature 'dreamletResult' seeErrors(obj) ## S4 method for signature 'dreamletProcessedData' seeErrors(obj) ## S4 method for signature 'vpDF' seeErrors(obj)
obj |
A |
tibble storing error text
library(muscat) library(SingleCellExperiment) data(example_sce) # create pseudobulk for each sample and cell cluster pb <- aggregateToPseudoBulk(example_sce, assay = "counts", cluster_id = "cluster_id", sample_id = "sample_id", verbose = FALSE ) # voom-style normalization res.proc <- processAssays(pb, ~group_id) # Differential expression analysis within each assay, # evaluated on the voom normalized data res.dl <- dreamlet(res.proc, ~group_id) # show errors # but none are reported res.err = seeErrors(res.dl)library(muscat) library(SingleCellExperiment) data(example_sce) # create pseudobulk for each sample and cell cluster pb <- aggregateToPseudoBulk(example_sce, assay = "counts", cluster_id = "cluster_id", sample_id = "sample_id", verbose = FALSE ) # voom-style normalization res.proc <- processAssays(pb, ~group_id) # Differential expression analysis within each assay, # evaluated on the voom normalized data res.dl <- dreamlet(res.proc, ~group_id) # show errors # but none are reported res.err = seeErrors(res.dl)
Show object
Show object
## S4 method for signature 'dreamletResult' show(object) ## S4 method for signature 'dreamletProcessedData' show(object)## S4 method for signature 'dreamletResult' show(object) ## S4 method for signature 'dreamletProcessedData' show(object)
object |
|
show data stored in object
Sort variance partition statistics
## S4 method for signature 'vpDF' sortCols( x, FUN = sum, decreasing = TRUE, last = c("Residuals", "Measurement.error"), ... )## S4 method for signature 'vpDF' sortCols( x, FUN = sum, decreasing = TRUE, last = c("Residuals", "Measurement.error"), ... )
x |
object returned by |
FUN |
function giving summary statistic to sort by. Defaults to sum |
decreasing |
logical. Should the sorting be increasing or decreasing? |
last |
columns to be placed on the right, regardless of values in these columns |
... |
other arguments to sort |
data.frame with columns sorted by mean value, with Residuals in last column
library(muscat) library(SingleCellExperiment) data(example_sce) # create pseudobulk for each sample and cell cluster pb <- aggregateToPseudoBulk(example_sce, assay = "counts", cluster_id = "cluster_id", sample_id = "sample_id", verbose = FALSE ) # voom-style normalization res.proc <- processAssays(pb, ~group_id) # variance partitioning analysis vp <- fitVarPart(res.proc, ~group_id) # Summarize variance fractions genome-wide for each cell type plotVarPart(sortCols(vp))library(muscat) library(SingleCellExperiment) data(example_sce) # create pseudobulk for each sample and cell cluster pb <- aggregateToPseudoBulk(example_sce, assay = "counts", cluster_id = "cluster_id", sample_id = "sample_id", verbose = FALSE ) # voom-style normalization res.proc <- processAssays(pb, ~group_id) # variance partitioning analysis vp <- fitVarPart(res.proc, ~group_id) # Summarize variance fractions genome-wide for each cell type plotVarPart(sortCols(vp))
Stack assays from pseudobulk to perform analysis across cell types
stackAssays(pb, assays = assayNames(pb))stackAssays(pb, assays = assayNames(pb))
pb |
pseudobulk |
assays |
array of assay names to include in analysis. Defaults to |
pseudobulk SingleCellExperiment cbind'ing expression values and rbind'ing colData. The column stackedAssay in colData() stores the assay information of the stacked data.
library(muscat) library(SingleCellExperiment) data(example_sce) # create pseudobulk for each sample and cell cluster pb <- aggregateToPseudoBulk(example_sce, assay = "counts", cluster_id = "cluster_id", sample_id = "sample_id", verbose = FALSE ) # Stack assays for joint analysis pb.stack <- stackAssays(pb) # voom-style normalization # assay (i.e. cell type) can now be included as a covariate res.proc <- processAssays(pb.stack, ~ group_id + stackedAssay) # variance partitioning analysis vp <- fitVarPart(res.proc, ~ group_id + stackedAssay) # Summarize variance fractions across cell types plotVarPart(sortCols(vp)) # Interaction analysis allows group_id # to have a different effect within each stacedAssay vp2 <- fitVarPart(res.proc, ~ group_id * stackedAssay) plotVarPart(sortCols(vp2)) # Interaction model using random effects form <- ~ (1 | group_id) + (1 | stackedAssay) + (1 | group_id:stackedAssay) #library(muscat) library(SingleCellExperiment) data(example_sce) # create pseudobulk for each sample and cell cluster pb <- aggregateToPseudoBulk(example_sce, assay = "counts", cluster_id = "cluster_id", sample_id = "sample_id", verbose = FALSE ) # Stack assays for joint analysis pb.stack <- stackAssays(pb) # voom-style normalization # assay (i.e. cell type) can now be included as a covariate res.proc <- processAssays(pb.stack, ~ group_id + stackedAssay) # variance partitioning analysis vp <- fitVarPart(res.proc, ~ group_id + stackedAssay) # Summarize variance fractions across cell types plotVarPart(sortCols(vp)) # Interaction analysis allows group_id # to have a different effect within each stacedAssay vp2 <- fitVarPart(res.proc, ~ group_id * stackedAssay) plotVarPart(sortCols(vp2)) # Interaction model using random effects form <- ~ (1 | group_id) + (1 | stackedAssay) + (1 | group_id:stackedAssay) #
Convert results table to matrix
tabToMatrix(tab, col, rn = "ID", cn = "assay")tabToMatrix(tab, col, rn = "ID", cn = "assay")
tab |
results table from |
col |
which column to extract |
rn |
column id storing rownames |
cn |
column id storing colnames |
matrix storing values of column col in rows defind by rn and columns defined by cn
Extract a table of the top-ranked genes from a dreamlet fit.
## S4 method for signature 'dreamletResult' topTable( fit, coef = NULL, number = 10, genelist = NULL, adjust.method = "BH", sort.by = "P", resort.by = NULL, p.value = 1, lfc = 0, confint = FALSE )## S4 method for signature 'dreamletResult' topTable( fit, coef = NULL, number = 10, genelist = NULL, adjust.method = "BH", sort.by = "P", resort.by = NULL, p.value = 1, lfc = 0, confint = FALSE )
fit |
dreamletResult object |
coef |
coef |
number |
number |
genelist |
genelist |
adjust.method |
adjust.method |
sort.by |
sort.by |
resort.by |
resort.by |
p.value |
p.value |
lfc |
lfc |
confint |
confint |
data.frame storing hypothesis test for each gene and cell type
limma::topTable(), variancePartition::topTable()
library(muscat) library(SingleCellExperiment) data(example_sce) # create pseudobulk for each sample and cell cluster pb <- aggregateToPseudoBulk(example_sce, assay = "counts", cluster_id = "cluster_id", sample_id = "sample_id", verbose = FALSE ) # voom-style normalization res.proc <- processAssays(pb, ~group_id) # Differential expression analysis within each assay, # evaluated on the voom normalized data res.dl <- dreamlet(res.proc, ~group_id) # show coefficients estimated for each cell type coefNames(res.dl) # extract results using limma-style syntax # combines all cell types together # adj.P.Val gives study-wide FDR topTable(res.dl, coef = "group_idstim", number = 3)library(muscat) library(SingleCellExperiment) data(example_sce) # create pseudobulk for each sample and cell cluster pb <- aggregateToPseudoBulk(example_sce, assay = "counts", cluster_id = "cluster_id", sample_id = "sample_id", verbose = FALSE ) # voom-style normalization res.proc <- processAssays(pb, ~group_id) # Differential expression analysis within each assay, # evaluated on the voom normalized data res.dl <- dreamlet(res.proc, ~group_id) # show coefficients estimated for each cell type coefNames(res.dl) # extract results using limma-style syntax # combines all cell types together # adj.P.Val gives study-wide FDR topTable(res.dl, coef = "group_idstim", number = 3)
Perform a competitive gene set analysis accounting for correlation between genes.
## S4 method for signature 'dreamletResult,GeneSetCollection' zenith_gsa( fit, geneSets, coefs, use.ranks = FALSE, n_genes_min = 10, inter.gene.cor = 0.01, progressbar = TRUE, ... ) ## S4 method for signature 'dreamlet_mash_result,GeneSetCollection' zenith_gsa( fit, geneSets, coefs, use.ranks = FALSE, n_genes_min = 10, inter.gene.cor = 0.01, progressbar = TRUE, ... )## S4 method for signature 'dreamletResult,GeneSetCollection' zenith_gsa( fit, geneSets, coefs, use.ranks = FALSE, n_genes_min = 10, inter.gene.cor = 0.01, progressbar = TRUE, ... ) ## S4 method for signature 'dreamlet_mash_result,GeneSetCollection' zenith_gsa( fit, geneSets, coefs, use.ranks = FALSE, n_genes_min = 10, inter.gene.cor = 0.01, progressbar = TRUE, ... )
fit |
results from |
geneSets |
|
coefs |
coefficients to test using |
use.ranks |
do a rank-based test |
n_genes_min |
minimum number of genes in a geneset |
inter.gene.cor |
if NA, estimate correlation from data. Otherwise, use specified value |
progressbar |
if TRUE, show progress bar |
... |
other arguments |
This code adapts the widely used camera() analysis (Wu and Smyth 2012) in the limma package (Ritchie et al. 2015) to the case of linear (mixed) models used by variancePartition::dream().
data.frame of results for each gene set and cell type
data.frame of results for each gene set and cell type
library(muscat) library(SingleCellExperiment) data(example_sce) # create pseudobulk for each sample and cell cluster pb <- aggregateToPseudoBulk(example_sce, assay = "counts", cluster_id = "cluster_id", sample_id = "sample_id", verbose = FALSE ) # voom-style normalization res.proc <- processAssays(pb, ~group_id) # Differential expression analysis within each assay, # evaluated on the voom normalized data res.dl <- dreamlet(res.proc, ~group_id) # Load Gene Ontology database # use gene 'SYMBOL', or 'ENSEMBL' id # use get_MSigDB() to load MSigDB library(zenith) go.gs <- get_GeneOntology("CC", to = "SYMBOL") # Run zenith gene set analysis on result of dreamlet res_zenith <- zenith_gsa(res.dl, go.gs, "group_idstim", progressbar = FALSE) # for each cell type select 3 genesets with largest t-statistic # and 1 geneset with the lowest # Grey boxes indicate the gene set could not be evaluted because # to few genes were represented plotZenithResults(res_zenith, 3, 1)library(muscat) library(SingleCellExperiment) data(example_sce) # create pseudobulk for each sample and cell cluster pb <- aggregateToPseudoBulk(example_sce, assay = "counts", cluster_id = "cluster_id", sample_id = "sample_id", verbose = FALSE ) # voom-style normalization res.proc <- processAssays(pb, ~group_id) # Differential expression analysis within each assay, # evaluated on the voom normalized data res.dl <- dreamlet(res.proc, ~group_id) # Load Gene Ontology database # use gene 'SYMBOL', or 'ENSEMBL' id # use get_MSigDB() to load MSigDB library(zenith) go.gs <- get_GeneOntology("CC", to = "SYMBOL") # Run zenith gene set analysis on result of dreamlet res_zenith <- zenith_gsa(res.dl, go.gs, "group_idstim", progressbar = FALSE) # for each cell type select 3 genesets with largest t-statistic # and 1 geneset with the lowest # Grey boxes indicate the gene set could not be evaluted because # to few genes were represented plotZenithResults(res_zenith, 3, 1)