Databases of liquid-liquid phase separation proteins and condensates

Protein liquid-liquid phase separation (LLPS) and the formation of biomolecular condensates are increasingly recognized as important in both cellular physiology and pathology. However, the computational databases developed to catalog LLPS-associated proteins differ significantly in scope, curation strategies, evidence models, and annotation granularity. Here, we compare the main LLPS databases, discuss what each resource captures well, where role definitions and evidence models diverge, which metadata remain limited, and how these differences affect data interpretation and reuse. We argue that future resources should prioritize interoperable vocabularies, explicit data origin, condition-aware annotations, and stronger integration with region-level, structural, and mutational information.