Prediction of protein secondary structure using Bayesian method and support vector machines

We propose a hybrid approach to predict the secondary structure of a protein from its amino acid sequence. Many existing techniques predict the secondary structure at each position of amino acid sequences based on a local window of residues. By combining the Bayesian method that avoids the problems of considering only a local neighborhood with Support Vector Machines (SVMs) which have optimal generalization, the new preditor achieves an accuracy of 70.9% when using the sevenfold cross validation on a database of 126 nonhomologous globular proteins. We show that it is possible to obtain a higher accuracy with the combined classifier than Bayesian classifier or Support Vector Machines, alone.