An annealing mutation operator in the genetic algorithms for RNA folding

Area of Science:

• Computational Biology
• Bioinformatics
• Algorithm Design

Background:

• RNA folding is a fundamental problem in molecular biology.
• Genetic algorithms (GA) are powerful tools for complex optimization problems like RNA folding.
• Efficient parallel computation is crucial for handling large biological sequences.

Purpose of the Study:

• To design and implement an efficient annealing mutation operator for genetic algorithms (GA) applied to RNA folding.
• To improve the convergence speed and efficiency of GA on parallel computing architectures like the MasPar MP-2.
• To develop a reliable termination technique for GA-based RNA folding simulations.

Main Methods:

• Design of a novel annealing mutation operator where mutation probability decreases hyperbolically with secondary structure size.
• Implementation of the operator on the MasPar MP-2 parallel computing system.
• Development of a GA termination technique based on the new mutation operator's behavior.

Main Results:

• The new mutation operator demonstrates efficient performance, especially for long RNA sequences (thousands of nucleotides).
• The operator facilitates faster convergence of free energy distributions across MasPar MP-2 processors.
• The developed termination technique effectively manages the GA process.

Conclusions:

• The designed annealing mutation operator significantly enhances the efficiency of GA for RNA folding on parallel architectures.
• This approach offers a scalable solution for analyzing large biological sequences.
• The study introduces an efficient and effective method for terminating GA simulations in bioinformatics.