An Algorithm-Based Investigation to Reveal the Differential Dynamics of Water inside Protein Cavities as a Function of Distance from Their Wall

Water dynamics within protein cavities influence cavity function, but the principles governing this remain poorly understood. Determining the number of water molecules and characterizing their behavior inside cavities are challenging despite biophysical advances. While molecular dynamics (MD) simulations offer insights into cavity-confined water, they are computationally expensive to study due to the short-lived nature and abundance of water-protein interactions. To address this, we present a multiprocessed algorithm that detects water inside protein cavities based on its distance from the cavity wall and analyzes its dynamics through residence times, orientational behavior, and diffusion. This software-independent method is capable of analyzing systems with millions of atoms in real time. Using molecular dynamics simulations, our algorithm distinguishes water inside the cavity from bulk water surrounding the protein with 1 Å resolution. We validate the method on multiple cavity-containing proteins and demonstrate its utility in understanding the biological role of water within the cavity of an archaeal group II chaperonin.