GEODES: Geometric Descriptors for the Assessment of Global and Local Flexibility of Proteins During Molecular Dynamics Simulation

Molecular dynamics simulations offer insights into macromolecular structures and functions through extensive time-series atomic data. Two widely used metrics for assessing these simulations are RMSD and RMSF. While RMSD measures conformational convergence, it suffers from degeneracy, where different conformations can produce identical values relative to a reference. RMSF indicates relative mobility but lacks temporal specificity. As both metrics provide only global perspectives, there is a pressing need for novel metrics to capture local flexibility with temporal resolution. We introduce GEODES, a novel complementary 3D geometrical descriptor approach, and compare its effectiveness with conventional analyses using RMSD and RMSF. Through molecular dynamics simulations of the vitamin D receptor trimeric complex, we demonstrate that GEODES significantly enhances the molecular dynamics analysis workflow, offering deeper insights into the structural dynamics and interactions of this complex. This innovative and versatile approach holds great promise for applications in drug discovery, structural biology, and bioinformatics. The GEODES Python3 script toolbox is available at https://github.com/rinnifox/GEODES