A Technique Based on Neural Network for predicting the Secondary Structure of Proteins

This paper presents a neural network approach for predicting the secondary structure of proteins from amino acid sequences. We have assumed a simple neural network with one input & output layer. Single hidden layer is also considered. Our neural network can be trained to predict the secondary structure of proteins by studying proteins with already known secondary structure by modifying the weight matrix function. Like other existing methods, our method is also an approximation method and its accuracy depends on the careful study and training of network. The significance of the method is its simplicity and applicability.