An efficient method based on hopfield neural network for rna secondary structure prediction

We present an improved method based on the Hopfield neural network for RNA secondary structure prediction in this study. The proposed method adjusts two parameters of the energy function in gradient ascent direction when the Hopfield neural network traps in a local minimum. The correction of the two parameters can increase the energy temporarily and help the network escape from the local minimum. The proposed algorithm was analyzed theoretically and evaluated experimentally through predicting RNA secondary structure. The simulation results on four RNA sequence show that the proposed algorithm performs better than others and has the ability to search the more stable RNA secondary structure for a RNA sequence.