New stochastic strategy to analyze helix folding

Area of Science:

• Computational biology
• Biophysics
• Structural bioinformatics

Background:

• Predicting polypeptide secondary structures is crucial for understanding protein function.
• Traditional methods like molecular dynamics and Monte Carlo can be computationally intensive.
• Conformational preferences, such as those defined by the Ramachandran map, guide protein folding.

Purpose of the Study:

• To develop an efficient stochastic strategy for searching polypeptide conformational space.
• To identify stable secondary structures, specifically alpha-helices, in peptides.
• To optimize the search using generalized simulated annealing (GSA) and Ramachandran map biases.

Main Methods:

• Utilizing the generalized simulated annealing (GSA) algorithm for conformational searching.
• Coupling a classical force field (THOR package) with GSA.
• Biasing peptide backbone angles (Phi x Psi) towards allowed regions of the Ramachandran map.
• Employing a continuum medium approach for energy calculations.

Main Results:

• Stable alpha-helix structures were obtained for polyalanines with 13 or more residues.
• The energy gap between global and local minima increases with polypeptide size.
• The GSA strategy proved significantly faster than traditional molecular dynamics or Monte Carlo methods.
• 2880 stochastic molecular optimizations were performed.

Conclusions:

• The GSA-based strategy offers an efficient and rapid method for predicting polypeptide secondary structures.
• Ramachandran map preferences effectively guide the search for stable alpha-helices.
• This approach optimizes the exploration of conformational space for peptide structure prediction.