BindProfX: Assessing Mutation-Induced Binding Affinity Change by Protein Interface Profiles with Pseudo-Counts.

Area of Science:

• Computational Biology
• Protein Engineering
• Biophysics

Background:

• Predicting mutation effects on protein-protein interaction (PPI) binding affinity is crucial for protein engineering.
• Current physical force field methods face challenges in accurately calculating mutation-induced binding free-energy changes (ΔΔG).

Purpose of the Study:

• To develop a novel, structure-based approach for calculating the impact of gene mutations on PPI binding affinity.
• To improve the accuracy and robustness of predicting binding free-energy changes.

Main Methods:

• Developed a structure-based profiling method using an interface alignment matrix to calculate ΔΔG.
• Incorporated three pseudo-counts to enhance the interface library.
• Combined the profile score with physical potentials (FoldX) for improved predictions.

Main Results:

• Achieved a ΔΔG prediction correlation of 0.68 for single-site mutations, up from 0.33 with previous methods.
• A combined profile and FoldX score yielded a correlation of 0.74.
• The profile score effectively captured the coupling effect of multiple mutations, maintaining strong correlation where physical potentials failed.

Conclusions:

• The proposed structure-based profile score offers a significant advancement in predicting mutation effects on protein binding affinity.
• This method is complementary to physical potentials and robust for complex mutations.
• The approach holds promise for protein engineering and understanding disease mechanisms involving altered protein interactions.