Compute a Deconvolution Dictionary From Subgroup Gene Programs — compute_deconvolution

For each subgroup, this function aggregates signature genes, ranks genes by correlation with subgroup abundance, performs GSEA, and renames subgroup labels using the top enriched pathway.

Usage

compute_deconvolution_dictionary(
  subgroups,
  expr,
  pathways = NULL,
  plot = FALSE
)

Arguments

subgroups: Output list from compute.deconvolution.analysis().
expr: Gene expression matrix with genes in rows and samples in columns.
pathways: Optional named list of pathways (pathway -> genes) used by fgsea. If NULL, Hallmark pathways are used.
plot: Logical; if TRUE, save GSEA summary plots in Results/.

Value

Updated subgroups list with renamed subgroup labels in "Deconvolution matrix" and "Deconvolution subgroups composition".