ABSTRACT: Artifi cial Neural Networks (ANNs) are simplifi ed models of the nervous system, in which neurons are consideredas simple processing units linked with weighted connections called synaptic effi cacies. These weights are graduallyadjusted according to a learning algorithm.Oxidoreductase any of a class of enzymes that catalyse oxidation–reductionreactions, i.e. they are involved in the transfer of hydrogen or electrons between molecules. They include the oxidasesand dehydrogenases.In this paper, an attempt has been made to develop a neural network-based method for predicting the secondarystructure of protein (Human Oxidoreductase family). The neural network has been trained using Bayesian RegularizationFeed-forward Backpropagation Neural Network Technique to predict the α-helix, β-sheet and coil regions of this proteinfamily. Feed-forward neural network have been trained by analyzing windows of 25 parameters for predicting the centralresidue of protein sequence. PSI-BLAST has been used for multiple-sequence alignment. SCOP and PDB databasehas been used for searching the primary and secondary structure of proteins and for training the data set. The methodcorrectly identifi es the secondary structure of Human Oxidoreductase family with more than 79% accuracy, which is wellabove any previously reported method.
Journal of Proteomics & Bioinformatics. 01/2010;