HydraMap

HydraMap predicts favorable hydration sites within protein binding pockets to inform protein-ligand binding energetics and desolvation energy estimation.

Key Features:

• Empirical Methodology: Uses statistical potentials derived from 10,987 crystal structures from the Protein Data Bank to quantify interactions between protein atoms and water molecules.
• Density Mapping and Clustering: Evaluates the probability of placing a water probe across the binding pocket to generate a density map that is processed with a clustering algorithm to identify explicit hydration sites.
• Validation and Performance: Validated against two external test sets with results comparable to 3D-RISM and WATsite while achieving a 30–1000× speed advantage over traditional methods.
• Desolvation Energy Estimation (DEWED): Estimates the desolvation energy associated with water replacement upon ligand binding, producing a DEWED term.
• Enhanced Scoring Functions: Integrates DEWED into ASP, ChemPLP, GoldScore, and X-Score, with reported improvements or maintenance of performance and notable accuracy gains for X-Score versus GB/SA-augmented variants.

Scientific Applications:

• Protein-ligand docking: Provides hydration-site and desolvation information to enhance scoring functions and improve docking prediction accuracy.
• Drug discovery and high-throughput screening: Enables rapid evaluation of hydration effects to support identification of potential therapeutic compounds in screening campaigns.
• Molecular modeling: Supplies explicit hydration maps and desolvation estimates that inform modeling of binding energetics and ligand design.

Methodology:

Derives statistical protein–water interaction potentials from 10,987 PDB crystal structures, computes water-probe placement probabilities to produce density maps, applies clustering to identify hydration sites, estimates DEWED desolvation terms for water replacement, integrates DEWED into ASP, ChemPLP, GoldScore and X-Score, and validates results against two external test sets compared to 3D-RISM and WATsite.