Predicting Thermodynamic Stability at Protein G Sites with Deleterious Mutations Using λ-Dynamics with Competitive...

Area of Science:

• Computational chemistry and molecular modeling.
• Protein engineering and bioinformatics.

Background:

• Free energy predictions are crucial for protein design and drug discovery.
• Alchemical free energy methods, particularly lambda-dynamics, offer high accuracy and improved computational efficiency.
• In silico site-saturation mutagenesis using lambda-dynamics enables the simulation of numerous mutations.

Purpose of the Study:

• To address the increased sampling demands of site-saturation mutagenesis, especially for deleterious mutations.
• To reintroduce and evaluate competitive screening with lambda-dynamics as a method to improve sampling efficiency.

Main Methods:

• Utilized lambda-dynamics simulations for in silico site-saturation mutagenesis.
• Implemented and compared competitive screening against traditional landscape flattening.
• Applied competitive screening to four surface and four buried sites in protein G.

Main Results:

• Competitive screening with lambda-dynamics demonstrated improved sampling, particularly for buried mutations.
• The method effectively addresses the need for increased sampling when characterizing multiple mutations.
• Successful application on protein G highlights its utility in molecular design.

Conclusions:

• Competitive screening with lambda-dynamics offers a viable strategy to enhance free energy predictions for protein design.
• The method provides new opportunities for exploring larger chemical spaces in molecular design.
• Improvements are particularly noted for challenging buried site mutations.