Integrating charge and non-charge patterning to predict Liquid-Liquid Phase Separation in IDPs

Liquid-liquid phase separation (LLPS) mediated by Intrinsically Disordered Proteins (IDP) has emerged as a fundamental process in cell biology. Predicting the temperature and density at which LLPS occurs is critical but remains challenging due to sequence effects—the order in which amino acids are linked in the protein backbone. Previous physics-based theory using random-phase-approximation combined with single chain renormalization (beyond Gaussian chain assumption) has been successful in predicting LLPS accounting for pH, salt, and sequence charge patterning, yet it neglects contributions from non-electrostatic interactions. We present an integrated framework that considers both charge and non-charge patterning to predict sequence-dependent LLPS. This approach is validated by comparing calculated phase diagrams with experimental data for various protein sequences, illustrating how sequence-specific interactions, mutations influence phase behavior beyond charge patterning.