Next-generation predictors of protein phase behavior

Biomolecular condensates formed through protein phase separation are critical for cellular organization and regulation. Recent years have seen rapid growth in computational methods predicting proteins' phase separation propensity and condensate localization, fueled by expanding datasets and advances in machine learning. Here, we review recent progress and limitations of state-of-the-art tools. Despite improvements, current models often fail to capture the complexity of phase separation, which depends on molecular interactions and contextual factors such as temperature, ionic strength, and macromolecular crowding. Encouragingly, new approaches are beginning to incorporate these biological variables, moving toward more physiologically relevant predictions. To accelerate progress, we advocate for stricter metadata standards and a coordinated, community-wide benchmarking of predictive tools to ensure robust and reproducible models for inference of protein phase behavior.