Accelerated Protein Folding Using Greedy-Proximal A.

Area of Science:

• Molecular Biophysics
• Biochemistry
• Computational Biology

Background:

• The protein folding problem is crucial in understanding diseases like Alzheimer's and Parkinson's.
• Molecular dynamics (MD) simulations are used to study protein folding but face timescale limitations.
• Existing methods to accelerate MD simulations often involve artificial energy biases.

Purpose of the Study:

• To introduce a novel, rational approach, Greedy-proximal A* (GPA*), for simulating protein folding pathways.
• To develop new protein structure comparison metrics based on contact map distance.
• To reduce computational time and improve the efficiency of protein folding simulations.

Main Methods:

• Developed and applied the Greedy-proximal A* (GPA*) algorithm, inspired by path-finding algorithms.
• Introduced novel contact map distance metrics for protein structure comparison.
• Tested GPA* on diverse protein structures: Trp-cage (TC5b), Protein G (1GB1), and Villin (1YRF).
• Compared GPA* performance against replica-exchange MD and steered MD.

Main Results:

• GPA* successfully identified shortest folding pathways for tested proteins.
• The algorithm significantly reduced computational time compared to standard MD methods.
• GPA* generated folding trajectories with minimal necessary motions, avoiding artificial energy bias.
• New contact map metrics proved effective in mitigating challenges with standard metrics.

Conclusions:

• Greedy-proximal A* (GPA*) offers an efficient and unbiased method for simulating protein folding.
• This approach enhances our understanding of protein folding dynamics and its relation to disease.
• GPA* represents a significant advancement in computational biophysics for studying protein folding pathways.