Despite the significant regulatory effects of protein post-translational modifications (PTMs) on liquid-liquid phase separation, specific data resources and predictors are lacking. Here, the authors report a manually curated database of PTM regulation (PTMPhaSe) and develop a graph neural network-based deep learning model (PhosLLPS) to predict functional phosphorylation sites regulating liquid-liquid phase separation, demonstrating their application to proteome-scale predictions.
- Xiaokun Hong
- Jiyang Lv
- Hai-Feng Chen