Predicting Saturation Concentrations of Phase-Separating Proteins via Thermodynamic Integration

Area of Science:

• Biophysics and Computational Biology
• Molecular and Cellular Biology

Background:

• Biomolecular condensates regulate cellular processes through phase separation.
• Saturation concentration (C_sat) is critical for condensate formation but challenging to simulate accurately.
• Current simulation methods lack statistical power for precise C_sat determination.

Purpose of the Study:

• To develop robust simulation techniques for accurate C_sat and phase diagram calculations.
• To enable direct comparison between simulation results and experimental data for biomolecular phase separation.
• To validate and improve coarse-grained models for predicting protein condensate behavior.

Main Methods:

• Introduced two independent thermodynamic integration (TI) schemes.
• Combined TI with Direct Coexistence simulations using the Mpipi-Recharged model.
• Compared TI methods with a machine-learning predictor for C_sat estimation.

Main Results:

• Accurately estimated C_sat for diverse proteins, including disease-associated and engineered variants.
• Enabled calculation of complete phase diagrams, revealing molecular mechanisms of phase separation.
• TI methods provide a physically grounded framework for validating biomolecular simulation models.

Conclusions:

• The novel TI schemes accurately determine C_sat and phase diagrams for biomolecular condensates.
• This approach bridges the gap between computational simulations and experimental observations.
• Enhanced validation of coarse-grained models for predicting protein phase behavior.