DrugScorePPI

DrugScorePPI predicts changes in binding free energy resulting from in silico alanine scanning of protein–protein interfaces.

Key Features:

• In silico alanine scanning: Performs alanine substitutions at protein–protein interfaces to evaluate effects of wild-type residue-to-Ala mutations on binding free energy.
• Knowledge-based scoring function: Employs a scoring function based on pair potentials derived from a dataset of 851 complex structures to evaluate inter-residue interactions.
• Experimental adaptation: Calibrates the scoring function against 309 experimental alanine scanning results to align predictions with empirical data.
• Output summaries: Produces summary tables and bar plots of binding free energy differences for each wild-type→Ala mutation.
• Structural mapping and visualization: Maps predicted binding free energy differences onto complex structures and provides visualization using Jmol.
• Validation: Performance was validated on an external test set of 22 alanine mutations in the Ras/RalGDS interface.

Scientific Applications:

• Identification of hotspot residues: Pinpoints residues that make major contributions to binding free energy and identifies interface hotspots.
• Guidance for biological experiments: Informs mutagenesis and interaction-mechanism studies by predicting energetic consequences of residue substitutions.
• Development of PPI modulators: Supports rational design of small molecules or peptides by highlighting key residues involved in protein–protein binding.

Methodology:

In silico alanine scanning uses a knowledge-based scoring function with pair potentials derived from 851 complex structures; the scoring function was adapted against 309 experimental alanine scanning results to predict binding free energy differences for wild-type→Ala mutations, generate summary tables and bar plots, map results onto structures and visualize with Jmol, and was validated on an external test set of 22 alanine mutations in the Ras/RalGDS interface.