Parts | Days | Selection | Search | Updates | Downloads | Help
BP: Fachverband Biologische Physik
BP 10: Computational Biophysics II
BP 10.2: Talk
Tuesday, March 19, 2024, 09:45–10:00, H 0112
Grand Canonical Simulations of Disordered Proteins — •Rodrigo F. Dillenburg1, Hao Ruan2, 3, Edward A. Lemke2, 3, and Martin Girard1 — 1Max Planck Institute for Polymer Research, Mainz, Germany — 2Institute of Molecular Biology, Mainz, Germany — 3Biocenter, Johannes Gutenberg University, Mainz, Germany
Investigations on Liquid-liquid phase separation (LLPS) have typically focused on intrinsically disordered proteins (IDPs), with theoretical support from polymer science. While great attention has been given to the study of large molecular condensates, little is known about non-deterministic smaller protein assemblies such as micelles. Such structures have been observed experimentally in artificially constructed sequences. We hypothesize that they could also arise in biologically relevant sequences. Coarse-grained force fields have provided an efficient framework for LLPS simulations with residue-level resolution and are remarkably accurate in reproducing phase diagrams of IDPs and the effects of residue mutations. Simulation methods designed for the study of molecular condensates must be modified to allow for simulations of microphases. The slab geometry devised to overcome slow diffusion times in highly dilute biological systems (~0.1% volume fraction), inhibits the formation of micelles. We implemented a Configurational Bias Monte Carlo algorithm based on the Rosenbluth-Rosenbluth method that allows for efficient LLPS simulations in a cubic simulation box and the investigation of microphase separation. We demonstrate the usefulness of this algorithm in the context of IDPs.
Keywords: Liquid-Liquid Phase separation; Intrinsically Disordered Proteins; Molecular Dynamics; Computational Methods; Monte Carlo simulations