Accelerated Hydration Site Localization and Thermodynamic Profiling

TL;DR

This paper introduces a rapid, accurate method using a geometric deep neural network to identify and analyze hydration sites around proteins, improving upon existing computational techniques by capturing complex interactions.

Contribution

The authors develop a novel deep learning approach trained on extensive molecular dynamics data for precise hydration site localization and thermodynamic profiling.

Findings

High accuracy confirmed on experimental data

Robustness demonstrated across multiple case studies

Outperforms traditional simplified models

Abstract

Water plays a fundamental role in the structure and function of proteins and other biomolecules. The thermodynamic profile of water molecules surrounding a protein are critical for ligand binding and recognition. Therefore, identifying the location and thermodynamic behavior of relevant water molecules is important for generating and optimizing lead compounds for affinity and selectivity to a given target. Computational methods have been developed to identify these hydration sites, but are largely limited to simplified models that fail to capture multi-body interactions, or dynamics-based methods that rely on extensive sampling. Here we present a method for fast and accurate localization and thermodynamic profiling of hydration sites for protein structures. The method is based on a geometric deep neural network trained on a large, novel dataset of explicit water molecular dynamics simulations. We confirm the accuracy and robustness of our model on experimental data and demonstrate it's utility on several case studies.