ACGA a Novel Biomimetic Hybrid Optimisation Algorithm Based on a HP Protein Visualizer: An Interpretable Web-Based...

Area of Science:

• Computational Biology
• Bioinformatics
• Molecular Modeling

Background:

• Protein structure prediction (PSP) is computationally challenging.
• Hydrophobic-polar (HP) lattice models simplify PSP by reducing conformational space.
• Existing metaheuristics struggle to find optimal protein structures.

Purpose of the Study:

• To develop a high-performing genetic algorithm (GA) for protein structure prediction.
• To improve the efficiency and accuracy of protein folding simulations.
• To introduce a novel visualization tool for enhanced interpretability.

Main Methods:

• Implementation of a novel all conformations biomimetic genetic algorithm (ACGA).
• Inclusion of specific translation operators alongside classical crossover and mutation.
• Development of an HP Protein Visualizer tool for molecular modeling and analysis.

Main Results:

• The ACGA increases the probability of finding optimal protein conformations with the lowest energy.
• The HP Protein Visualizer provides interpretability and insight into the optimization process.
• The developed tool simulates protein folding using modern web technologies.

Conclusions:

• The ACGA offers a promising approach to overcoming challenges in protein structure prediction.
• The integrated visualizer enhances the understanding and analysis of protein folding.
• This work provides a valuable molecular modeling tool for studying protein structures.