Abstract

The formation of biomolecular condensates via liquid-liquid phase separation (LLPS) is one of the primary mechanisms used by cells for spatiotemporal organisation. Importantly, biomolecular phase separation has been directly linked to many biological functions such as heterochromatin organisation and transcription, DNA repair, and ribonucleoprotein formation, as well as dysfunction, via the formation of pathological aggregates. Although experimental techniques are essential for studying the bulk phase behaviour of biomolecules, computer simulations are able to provide more close-up views and intricate molecular details of this phenomenon. Over the last few years, the scientific community has made great progress in developing computational models to study LLPS , which has helped us link the biomolecular sequences to their macroscopic phase behaviour and decipher the physicochemical determinants of these transitions. I will present Mpipi, a multiscale coarse-grained model built upon bioinformatics data and atomistic calculations. ‘The model is designed to capture the dominant role of π–π and cation–π interactions, and the stronger contribution of arginine vs. lysine in LLPS and to carefully balance out the contribution of each amino acid. Using the Mpipi model, we compute phase diagrams for a series of well-studied proteins with near quantitative accuracy. I will describe the process of parameterising the force field, the comparison against other coarse-grained models, and the validation of the Mpipi model against existing experimental and computational studies.