A Data-Driven Hydrophobicity Scale for Predicting Liquid-Liquid Phase Separation of Proteins

Area of Science:

• Biophysics
• Molecular Biology
• Cell Biology

Background:

• Membraneless organelles form via liquid-liquid phase separation (LLPS).
• Understanding protein interactions driving LLPS is crucial for studying these structures.
• Existing hydrophobicity scales are not optimized for LLPS prediction.

Purpose of the Study:

• To develop a novel, data-driven hydrophobicity scale for proteins undergoing LLPS.
• To improve the prediction of LLPS in biological macromolecules.
• To better understand the molecular drivers of membraneless organelle formation.

Main Methods:

• Coarse-grained molecular dynamics simulations.
• Development of a new hydrophobicity scale using the force-balance method.
• Testing the scale on unfolded, intrinsically disordered, and phase-separating proteins (PSPs).

Main Results:

• The new hydrophobicity scale accurately predicts LLPS at physiological conditions.
• It outperforms existing scales in predicting LLPS using simulations.
• The scale highlights the significance of π-π interactions in driving LLPS.

Conclusions:

• The developed hydrophobicity scale offers a compact description of protein-protein interactions relevant to LLPS.
• This scale can enhance models for studying PSPs and their interactions with other cellular components.
• It provides a valuable tool for investigating the biophysics of membraneless organelles.