RNA secondary structure prediction algorithm based on combinatorial optimization algorithm and SVMs method

摘要

A new RNA secondary structure prediction algorithm that can predict pseudoknots is proposed, it combines the stem-loop combinatorial optimization algorithm and SVMs(Support Vector Machines, SVMs) method. The algorithm firstly finds out the optimal stem-loop structure and suboptimum structures based on dynamic neighbor topology particle swarm optimization algorithm, and then puts these loops into SVMs. The output from SVMs can decide whether there exist a pseudoknot. The experimental results demonstrate the superiority of our algorithm over the other methods in terms of solution quality and convergence rates.