Skip to contents

ExportCytoscape

ExportCytoscape.Rd

Export an mmo object as a node table and edge table for visualisation in Cytoscape. Edges are derived from a stored feature dissimilarity matrix. Nodes carry abundance statistics and any annotation columns present in mmo$sirius_annot and mmo$feature_info.

Usage

ExportCytoscape(
  mmo,
  distance = "dreams",
  outprefix = "cytoscape_export",
  sim_threshold = 0.7,
  top_k = NULL,
  group_col = "group",
  sample_col = "sample"
)

Arguments

mmo

mmo object. Must contain mmo$feature_data and mmo$metadata. mmo$feature_info and mmo$sirius_annot are used when present.

distance

Name of the dissimilarity matrix to use for edges (default: "dreams"). Passed to GetDistanceMat().

outprefix

File path prefix for output CSVs (default: "cytoscape_export").

sim_threshold

Minimum pairwise similarity (= 1 - dissimilarity) for an edge to be retained (default: 0.7). Set to 0 to disable.

top_k

Integer or NULL. If not NULL, each node retains only its top_k most similar neighbours after threshold filtering (default: NULL = no k limit).

group_col

Metadata column defining groups for per-group mean abundance columns (default: "group").

sample_col

Metadata column matching sample IDs to feature columns (default: "sample").

Value

Invisibly returns a named list: nodes (data.frame), edges (data.frame), nodes_path, edges_path (file paths).

Details

Edge filtering: Two complementary filters control which edges are retained. Both are applied together (a pair must pass both to be included):

  • sim_threshold – drop all edges with similarity below this value. Higher values produce sparser, higher-confidence networks.

  • top_k – for each node, retain only its top_k most similar neighbours (by similarity). NULL keeps all edges that pass the threshold. This is useful for preventing highly-connected hub nodes from dominating the layout.

A warning is issued when the retained edge count exceeds 50 000, as large networks can be slow to render in Cytoscape.

Node table columns: The node table always includes id, prevalence (proportion of samples with detected abundance), and one mean_<group> column per group. Any columns present in mmo$feature_info and mmo$sirius_annot are appended automatically – the function does not assume specific column names. Column names are sanitised to replace characters that cause problems in Cytoscape (spaces, #, :).

Loading in Cytoscape:

  1. File -> Import -> Network from File -> select _edges.csv. Map source as Source Node, target as Target Node, similarity as Edge Attribute.

  2. File -> Import -> Table from File -> select _nodes.csv. Map id as the Key column matching node names.

  3. In the Style panel, map Fill Color to any annotation column (e.g. NPC_pathway) to color nodes by compound class.

Examples

if (FALSE) {
# Default: similarity >= 0.7
ExportCytoscape(mmo, distance = "dreams",
                outprefix = "output/network")

# Sparser network: top 5 neighbours per node, similarity >= 0.6
ExportCytoscape(mmo, distance = "dreams",
                outprefix = "output/network_k5",
                sim_threshold = 0.6, top_k = 5)
}