An effective evolutionary algorithm for protein folding on 3D FCC HP model by lattice rotation and generalized move...

Area of Science:

• Biochemistry
• Molecular Biology
• Biophysics

Background:

• Protein tertiary structure dictates function, making its prediction from amino acid sequence a key challenge.
• The Hydrophobic-Polar (HP) lattice model is a simplified ab initio method for forecasting protein structures.
• This model aids in understanding protein folding principles and structure prediction.

Purpose of the Study:

• To present an improved evolutionary algorithm (EA) for the protein folding problem.
• To enhance EA robustness, implementation ease, and energy function handling.
• To optimize protein structure prediction on the 3D FCC lattice HP model.

Main Methods:

• Developed an improved evolutionary algorithm (EA) for protein folding.
• Utilized the 3D FCC lattice HP model for simulations.
• Integrated three local search techniques: lattice rotation (crossover), K-site move (mutation), and generalized pull move.

Main Results:

• Experimental validation on established datasets.
• Demonstrated the approach's ability to find optimal conformations previously undiscovered by other EA-based methods.
• The improved EA outperforms existing approaches in identifying optimal protein structures.

Conclusions:

• Investigated 3D FCC lattice geometric properties and developed novel local search techniques.
• EA-based approaches are robust and adaptable to various energy functions.
• Enhanced local searches significantly improve EA effectiveness for optimal conformation discovery on the 3D FCC HP model, with potential integration into other search methods.