This tool analyzes how frequently pairs of interactions (e.g., H-bond A–B and Salt Bridge C–D) co-occur in the same simulation frames. It visualizes these pairwise co-occurrence frequencies as a matrix-style percentage heatmap.

• Objective: Identify patterns of correlated interactions by quantifying simultaneous presence across trajectory frames.
• Process:
  • Loads interaction presence matrices generated during the ▶️ (Run First) Extract Pairwise Interaction from Trajectory step.
  • Filters out interactions below a frequency threshold (e.g., present in <40% of frames).
  • Computes pairwise co-occurrence percentages between all interactions.
  • Visualizes results using a symmetric heatmap and a colorbar.

---

• Requires a loaded trajectory.
• Requires interaction CSVs generated using the same sel1_name and sel2_name from the extraction step.

• Selection Labels:

  • sel1_name, sel2_name: Labels used to locate and match interaction data files.

• Interaction Mode:

  • interaction_mode: Scope of interaction analysis.
    • interchain + intrachain: All interactions.
    • interchain: Between different chains.
    • intrachain: Within the same chain.

• Interaction Type Filter:

  • interaction_types: Specify which interaction types to include in the analysis. Choose from:
    • All, H-Bonds, Salt Bridges, Hydrophobic, π-Stacking, Cation–π.

• Contact Filtering:

  • pct_hide_threshold: Only include interactions present in at least this percentage of frames.

• Parallel Processing:

  • num_threads: Number of CPU threads to use for co-occurrence calculation. Use -1 to use all available threads.

---

• Co-occurrence Heatmap:

  • Matrix where each cell shows the percentage of frames in which both interactions in a pair are present.
  • Diagonal cells are always set to 100%.
  • Interactions are labeled by residue pairs (e.g., ARG123–GLU45).

• Colorbar:

  • Gradient scale from 0% to 100% co-occurrence.
  • Independent from the main heatmap for modular layout.

• Saved Files:

  • interaction_co_occurrence_heatmap_<sel1>_<sel2>.png: Heatmap image.
  • interaction_co_occurrence_<sel1>_<sel2>.csv: Raw co-occurrence matrix (in %).

• Console Output:

  • Lists excluded interactions below the threshold.
  • Shows progress during multiprocessing steps.
  • Prints legend when all interaction types are included.

---

• High Co-occurrence:

  • High values (close to 100%) indicate strong synchronization or cooperative presence of two interactions.

• Low Co-occurrence:

  • Low percentages suggest mutually exclusive, transient, or unrelated interactions.

• Thresholding Effect:

  • The pct_hide_threshold setting directly controls the number of interactions visualized. Raising it filters out noise; lowering it increases detail.

• Sidechain Indicator:

  • Residue labels with * indicate sidechain-originating interactions (inherited from extraction data).

---

• Scenario: Determine if certain ligand–residue contacts frequently appear together.
• Selections:
  • sel1 = 'protein'
  • sel2 = 'resname LIG'
• Observation: Persistent co-occurrence of ligand-sidechain contacts suggests cooperative binding pocket regions.

• Scenario: Explore whether contacts in different protein regions fluctuate together.
• Selections:
  • sel1 = 'resid 10-50'
  • sel2 = 'resid 150-190'
• Observation: Non-adjacent contacts with high co-occurrence may indicate long-range coupling.

---

• For Full Interaction Scope:

  • Set interaction_types = 'All' to compute comprehensive co-occurrence across interaction classes.

• Legend Readability:

  • When showing all interaction types, refer to the printed legend for abbreviation meanings.

• Matrix Size Handling:

  • Increase fig_width_factor or fig_height_factor to avoid overcrowded axes for large datasets.

---