Generate Node Attributes for a Complex Network

Creates a detailed attribute table for each complex, emphasizing functional specificity for network visualization.

Usage

generateNodeAttributes(
  complexes,
  enrichments,
  geneSetDb = NULL,
  similarityMethod = "jaccard",
  verbose = TRUE
)

Arguments

complexes

A named list of protein complexes.

enrichments

A named list of enrichment results. Must contain 'Fold' column for specificity weighting.

geneSetDb

Optional named list of gene sets for semantic clustering.

similarityMethod

Distance method for clustering ("jaccard", "overlap", etc.). Defaults to "jaccard" to penalize size differences.

verbose

Logical.

Value

A `tibble` with complex attributes.

Details

This function performs several steps to generate rich node attributes: 1. It aggregates all enriched terms from the input `enrichments` list. 2. It calculates a **Specificity Score** for each term using the formula: `-log10(p.adjust) * log2(Fold)`. This ensures that specific, high-fold enrichment terms are prioritized over broad, generic terms. 3. A term-complex matrix is built, and terms are clustered. The clustering uses **Average Linkage** and forces a higher number of clusters to preserve functional diversity (avoiding "monochromatic" maps). 4. A unique color is assigned to each functional domain. 5. For each complex, the **Primary Functional Domain** is assigned to the enriched term with the highest Specificity Score. 6. A unique "blended" color is calculated by mixing domain colors, weighted by the Specificity Score.

Author

Qingzhou Zhang <zqzneptune@hotmail.com>