enriched_features: Tidy CharacteristicFeatures() output for plotting

In MosaiClusteR: An Umbrella Framework for Multi-Source and Multi-Omics Clustering

Tidy CharacteristicFeatures() output for plotting

Description

Turns the output of CharacteristicFeatures() into a per-cluster list of compact tables (term / pvalue / Count) ready for plot_pathways(). For each cluster it keeps the characteristic features that pass a significance cut and, by default, only those that are enriched in the cluster (more frequent among its objects than in the remaining objects) rather than depleted. Empty clusters are dropped.

Usage

enriched_features(
  char,
  data,
  source = 1,
  method = 1,
  sign = 0.05,
  enriched_only = TRUE,
  clean_names = TRUE
)

Arguments

char	Output of CharacteristicFeatures().
data	The binary data matrix whose features were tested (objects in rows, features in columns) - e.g. the target matrix. Used to set the enrichment direction and the per-cluster feature counts.
source	Which tested data source to use within each cluster, by name (e.g. "Target") or index. Default: the first source.
method	Which clustering result in char to use, by name or index. Default 1.
sign	Keep features with adjusted p-value below this. Default 0.05.
enriched_only	If TRUE (default) keep only features more frequent in the cluster than in the rest of the objects; if FALSE keep every significant characteristic feature (enriched or depleted).
clean_names	If TRUE (default) tidy the feature names for display - reversing the make.names() mangling common in binary descriptor matrices (drop a leading X before a digit, turn ./⁠_⁠ runs into spaces), e.g. "X3.oxo.5.alpha_steroid" becomes "3 oxo 5 alpha steroid". Set FALSE to keep the raw column names.

Value

A named list - one element per non-empty cluster - of data frames with columns term, pvalue (adjusted p-value) and Count (number of the cluster's objects carrying the feature), suitable for plot_pathways().

See Also

CharacteristicFeatures(), plot_pathways()

Examples

## Not run: 
Char <- CharacteristicFeatures(List = list(Weighted = sol),
                               binData = list(Target = targetMat),
                               datanames = "Target", nrclusters = 4)
tabs <- enriched_features(Char, targetMat)
plot_pathways(tabs, style = "circular")

## End(Not run)

MosaiClusteR documentation built on Oct. 25, 2026, 1:08 a.m.