Map study factors to TCGA meta-programs — map_factors_to

For each study NMF factor, scores it against each TCGA meta-program by computing the mean NES of the meta-program's Hallmarks in that factor. The meta-program with the highest positive mean NES is the best match. If the meta-program reference has a TME_subtype column (see annotate_metaprograms_TME()), it is appended to the output.

Usage

map_factors_to_metaprograms(
  gsea_study,
  cancer_type,
  mp_file = NULL,
  plot = TRUE,
  file_name = NULL
)

Arguments

gsea_study: Output from compute_factor_gsea() on study cohort.
cancer_type: Character. TCGA cancer type abbreviation identifying which pre-built TCGA meta-program reference shipped with the package to use. Available: "blca", "luad", "skcm". Ignored if mp_file is given.
mp_file: Optional. Either a meta-program data frame (as returned by derive_meta_programs() / annotate_metaprograms_TME()) or the path to an RData file containing such an object named meta_programs. If NULL, the pre-built reference for cancer_type is used.
plot: Logical. If TRUE (default), saves a barplot of factor-to-meta-program scores to Results/Factor_MP_mapping_<file_name>.pdf.
file_name: Optional character suffix for saving output plots.

Value

A list with:

factor_mapping: Data frame with one row per study factor and columns best_MP, factor, best_score, all_scores (all meta-program scores as "MP:score" pairs) and, if available, TME_subtype.
reference: The meta-program reference data frame used.