Skip to contents

Differential expression analysis between two groups of cells using limma

Source: R/RunDE.R
RunLimmaDE.Rd

Performs differential expression analysis between two groups of cells using the limma linear modeling framework. Supports optional observation-level weights (e.g., spatial weights) and filters genes by minimum expression threshold across groups.

Usage

runLimmaDE(
  exp_mat = NULL,
  cells_reference = NULL,
  cells_target = NULL,
  weights = NULL,
  adj_p.value = 0.05,
  min.pct = 0
)

Arguments

exp_mat

A normalized gene expression matrix (genes x cells), either a matrix or dgCMatrix. Typically log-normalized counts, e.g., from a Seurat object.

cells_reference

A character vector of cell IDs to use as the reference (baseline) group.

cells_target

A character vector of cell IDs to use as the target (comparison) group.

weights

Optional numeric vector of observation-level weights. Must be named with cell IDs and match the length of cells_reference + cells_target.

adj_p.value

Adjusted p-value threshold for reporting differentially expressed genes. Default is 0.05.

min.pct

Minimum proportion of cells expressing the gene in either group (values between 0 and 1). Genes not meeting this threshold are excluded before testing. Default is 0.

Value

A data frame with differentially expressed genes, sorted by absolute log fold change. Includes columns:

logFC

Log2 fold change of expression (target vs. reference)

AveExpr

Average expression across both groups

t, P.Value, adj.P.Val, B

Statistical results from limma

pct.reference

Proportion of reference cells expressing the gene

pct.target

Proportion of target cells expressing the gene

gene

Gene name (rownames from exp_mat)

Examples

# Load coordinates and log-normalized expression data
coords <- readRDS(system.file("extdata", "MouseBrainCoords.rds",
    package = "SpNeigh"
))
logNorm_expr <- readRDS(system.file("extdata", "LogNormExpr.rds",
    package = "SpNeigh"
))

# Subset cells from cluster 0 and 2
cells_ref <- subset(coords, cluster == 0)$cell
cells_tar <- subset(coords, cluster == 2)$cell

# Run differential expression with minimum expression threshold
tab <- runLimmaDE(
    exp_mat = logNorm_expr,
    cells_reference = cells_ref,
    cells_target = cells_tar,
    min.pct = 0.25
)

head(tab[, c("gene", "logFC", "adj.P.Val", "pct.reference", "pct.target")])
#>          gene     logFC adj.P.Val pct.reference pct.target
#> Gjc3     Gjc3  4.502163         0    0.25993020  0.9000000
#> Sox10   Sox10  4.147574         0    0.08259930  0.8078571
#> Fn1       Fn1 -3.769727         0    0.79707495  0.1539286
#> Opalin Opalin  3.694963         0    0.06847266  0.6757143
#> Cldn5   Cldn5 -3.472197         0    0.68140269  0.0900000
#> Ly6a     Ly6a -3.290134         0    0.77829483  0.2583929