Prediction and characterization of liquid-liquid phase separation of minimalistic peptides

Liquid-liquid phase separation (LLPS) of proteins mediates the assembly biomolecular condensates involved in physiological and pathological processes. Identifying minimalistic building blocks sequence determinant protein is an urgent need but remains challenging partially due to lack methodologies characterize behavior. Here, we demonstrate computational tools efficiently quantify microscopic fluidity liquid temperature-dependent diagram peptides. We comprehensively explore LLPS abilities all 400 dipeptide combinations coded amino acids observe occurrences spontaneous three categories dipeptides. Our predictions are validated by turbidity assays differential interference contrast (DIC) microscopy. that dipeptides, minimal complete, possess multivalent interactions sufficient for LLPS, suggesting a general property peptides proteins, independent their length. This study paves way prediction characterization peptide behavior at molecular level. • A framework developed predict propensity Three dipeptides identified have high propensities work reveals The mechanism provided QW condensates. By exploring Tang et al. LLPS. previous studies suggest proteins.