Skip to contents

Compute cell group scores from deconvolution and TF module network

cell.groups.computation.Rd

This function identifies cell groups based on dendrogram cuts, computes composite scores for each group using deconvolution features and TF module networks, and optionally exports the results.

Usage

cell.groups.computation(
  deconvolution,
  cell.dendrograms,
  tfs.module.network,
  batch = NULL,
  return = T,
  pval = 0.05,
  n_perm = 999,
  dendrogram_file = NULL,
  return_dendrogram = FALSE
)

Arguments

deconvolution

A data frame with deconvolution features (samples as rows, cell-type features as columns). This is usually the first element returned by multideconv::compute.deconvolution.analysis().

cell.dendrograms

A named list of dendrogram objects, each corresponding to a TF module, typically returned by identify.cell.groups().

tfs.module.network

A list containing network information of transcription factor (TF) modules, as obtained from compute.WTCNA(). It should contain at least one element with TF module membership or connectivity.

batch

Optional vector indicating batch assignment for samples.

return

Logical; if TRUE (default), writes CSV files with cell group compositions and scores to the "Results/" folder.

pval

Numeric. P-value threshold for statistical tests. Default is 0.05.

n_perm

Integer. Number of permutations for significance testing. Default is 999.

dendrogram_file

Optional character. File path to save dendrogram plot.

return_dendrogram

Logical. If TRUE, saves a PDF of the colored cell-group dendrograms to Results/Dendrogram_color_clusters_<dendrogram_file>.pdf (only when dendrogram_file is set). Default FALSE.

Value

A list of three elements:

scores

A data frame with the composite scores of all identified cell groups across samples.

composition

A list of vectors indicating the composition (original features) of each cell group.

loadings

A list of CCA projection parameters (xcoef, train_means, train_sds) for each cell group.

If return=TRUE, two CSV files will be created:

  • Results/Cell.groups.composition.csv: A table showing the composition of each cell group.

  • Results/Cell.groups.scores.csv: A matrix of cell group scores across samples.

Examples

if (FALSE) { # \dontrun{
deconv_results <- multideconv::compute.deconvolution.analysis(...)
tf_network <- compute.WTCNA(...)
dendrograms <- identify.cell.groups(...)

cell.groups <- cell.groups.computation(
  deconvolution = deconv_results[[1]],
  cell.dendrograms = dendrograms,
  tfs.module.network = tf_network,
  return = TRUE
)
} # }