A hybrid approach to protein folding problem integrating constraint programming with local search

Area of Science:

• Computational Biology
• Bioinformatics
• Biophysics

Background:

• The protein folding problem is a significant challenge in computational biology.
• Simplified lattice models and energy functions aid in addressing its computational complexity.
• Heuristic search and constraint programming are common approaches.

Purpose of the Study:

• To introduce a novel hybrid approach for protein folding simulation.
• To integrate constraint programming within a local search framework.
• To improve the efficiency and accuracy of protein folding predictions.

Main Methods:

• Utilized a face-centered-cubic lattice model and a 20 amino acid pairwise interaction energy function.
• Applied constraint programming to generate neighboring conformations for local search.
• Integrated constraint programming with local search for protein structure prediction.

Main Results:

• The hybrid approach demonstrated substantial improvements in final energy values.
• Achieved better results compared to pure constraint programming and pure local search methods.
• Maintained acceptable runtimes for small and medium-sized proteins.

Conclusions:

• Hybrid techniques combining constraint programming and local search offer improved solutions for protein folding.
• This approach yields better quality protein structures in acceptable timeframes.
• Encourages the adoption of hybrid methods for complex optimization problems in computational biology.