PPI-Scanner: Predicting Protein–Protein Interactions

PPI-Scanner predicts protein–protein interactions (PPIs) based on 3D structures. Given two protein files (e.g., PDB), it analyzes their surfaces to determine if they are likely to bind.

Using a Siamese graph neural network, the tool extracts surface patches, encodes them with shape and charge features, and compares them to assess interaction likelihood.

Results include a binary prediction (binding/non-binding) and highlight key contact residues on both proteins.