dreamlet: Differential expression for each assay

In GabrielHoffman/dreamlet: Scalable differential expression analysis of single cell transcriptomics datasets with complex study designs

Differential expression for each assay

Description

Perform differential expression for each assay using linear (mixed) models

Usage

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,
  ...
)

Arguments

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 assayNames(x)
contrasts	character vector specifying contrasts specifying linear combinations of fixed effects to test. This is fed into makeContrastsDream( formula, data, contrasts=contrasts)
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 eBayes be used on result? (defualt: TRUE)
...	other arguments passed to dream

Details

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.

Value

Object of class dreamletResult storing results for each cell type

See Also

variancePartition::dream(), variancePartition::makeContrastsDream()

Examples

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)

GabrielHoffman/dreamlet documentation built on Oct. 29, 2024, 4:04 a.m.