Proteins 2D Structure Prediction from 1D Sequence by Signal Processing and Soft Computing Methods

Article history: Received Accepted Available online 10 Jan. 2015 20 Feb. 2015 25 Feb. 2015 Research in protein structure and function is one of the most important research areas in modern bioinformatics and computational biology. The structure of a protein is related to its function. The information necessary for protein folding resides completely within the primary structure. The development of rapid methods of DNA sequencing coupled with the straightforward translation of the genetic code into protein sequences has amplified the urgent need for automated methods of interpreting these onedimensional(1D), linear sequences in terms of two-dimensional(2D) structure. There is a considerable room for improvement. In this paper an effective feature extraction method based on discrete wavelet transform (DWT) to detect informative proteins. Support vector machine, Multilayered perceptron (MLP) and radial basis function (RBF) neural network classifiers are used to efficiently to predict the protein secondary structure which aims to classify the three types of α-helix, β-sheet and C-coil. Effective numerical representation based on physico-chemical parameters such as EIIP, polarizabilty, dipole moment and alpha induces the prediction more accurately. The potential of the proposed approach is evaluated through an exhaustive study by benchmark non-redundant dataset and a prediction accuracy of 95% is achieved. © 2015 International Journal of Advanced Research in Science and Technology (IJARST). All rights reserved.