An ant colony optimisation algorithm for the 2D and 3D hydrophobic polar protein folding problem

Area of Science:

• Computational molecular biology
• Biochemical physics
• Bioinformatics

Background:

• The protein folding problem is a central challenge in computational molecular biology and biochemical physics.
• Existing optimization methods for protein folding include Monte Carlo, Evolutionary Algorithms, and Tabu Search.
• This study introduces an Ant Colony Optimization (ACO) algorithm for the 2D and 3D hydrophobic-polar (HP) model.

Purpose of the Study:

• To present an improved ACO algorithm for the 2D HP model and extend it to the 3D HP model.
• To evaluate the performance of the new algorithm, ACO-HPPFP-3, against state-of-the-art methods.
• To assess the algorithm's ability to find diverse native conformations.

Main Methods:

• Development and application of an Ant Colony Optimization (ACO) algorithm tailored for protein folding.
• Testing the algorithm on the 2D and 3D hydrophobic-polar (HP) models.
• Comparison with existing state-of-the-art algorithms for protein structure prediction.

Main Results:

• The ACO-HPPFP-3 algorithm demonstrates superior performance on sequences with internal structural nuclei compared to end-located nuclei.
• The algorithm generally identifies a more diverse set of native protein conformations.
• Empirical results show favorable comparison with specialized methods for 2D and 3D HP models.

Conclusions:

• Ant Colony Optimization (ACO) is a promising approach for the 2D and 3D HP protein folding problem.
• While scaling may be a limitation with sequence length, the algorithm excels at finding diverse native states.
• Further development of ACO algorithms for more complex protein models holds significant potential.