An Evolutionary Profile Guided Greedy Parallel Replica-Exchange Monte Carlo Search Algorithm for Rapid Convergence in...

Area of Science:

• Computational biology
• Protein engineering
• Biophysics

Background:

• Protein design, or inverse protein folding, is crucial for creating novel proteins but remains an NP-hard problem.
• Existing heuristic approaches often focus on scoring functions, with room for improvement in search efficiency.

Purpose of the Study:

• To develop an efficient algorithm for protein design that addresses the NP-hard nature of the inverse protein folding problem.
• To enhance the speed and accuracy of identifying protein sequences that fold into desired structures.

Main Methods:

• A replica-exchange Monte Carlo (REMC) search algorithm was developed, incorporating a greedy strategy for faster convergence.
• An evolutionary profile, derived from structural homologs, was used to guide the REMC search.
• Stagnant simulation trajectories were terminated early to optimize computational resources.

Main Results:

• The algorithm achieved an average root-mean-square deviation of 1.21Å on a benchmark dataset, indicating high accuracy in protein modeling.
• A 6.16-fold overall speedup was observed compared to existing methods.
• Molecular Dynamics simulations showed that four out of five designed proteins exhibited comparable or improved stability relative to their target structures.

Conclusions:

• The proposed REMC algorithm with an evolutionary profile offers an efficient and effective solution for the protein design problem.
• The method demonstrates significant speedup and high accuracy in generating stable protein sequences.
• This approach advances the field of protein engineering and computational biology.