Force fields for simulating intrinsically disordered proteins: Assessing conformational sampling and structural dynamics

Abstract

Intrinsically disordered proteins (IDPs) lack a defined three-dimensional (3-D) structure but play crucial roles in biological pathways. They exist as conformational ensembles and experimental methods struggle to capture their dynamic nature, making molecular dynamics (MD) simulations a valuable tool. However, force field accuracy and sampling algorithms limit simulation fidelity. Most protein force fields are too stable to accurately model unstructured proteins like IDPs. Empirical force fields-based computer simulations are increasingly used to study the biophysics of disordered proteins, with the choice of force field significantly influencing simulation outcomes for studying the conformational ensemble of IDPs. This study evaluates three AMBER force fields (ff99SBildn-TIP3P, ff99SB-TIP3P, and ff19SB-OPC) for simulating an IDP (Histatin 5) and a partially folded protein (Trp-cage). Extensive MD simulations compared the structural dynamics and conformational sampling across all the force fields for these two systems. The results show ff99SBildn-TIP3P as the most balanced force field, efficiently sampling ordered and disordered regions in these proteins. We evaluated the performance of the force fields with enhanced sampling metrices including RMSD, RMSF, Rg and SASA. Our results reveal ff99SBildn-TIP3P model better samples the disorder regions in Histatin 5 than the other force fields. This study highlights the importance of understanding force field strengths and limitations for IDP simulations. By selecting suitable force fields, researchers can better simulate the IDPs and understand their complex behavior.