Protein fold classification is an important subject in computational biology and 8 a compelling work from the point of machine learning. To deal with such a challenging 9 problem, in this study, we propose a solution method for the classification of protein folds 10 using Grow and Learn neural network (GAL) together with one-versus-others (OvO) 11 method. To classify the most common 27 protein folds, 125 dimensional data, constituted 12 by the physicochemical properties of amino acids, are

more »

... The study is conducted on a 13 database including 694 proteins: 311 of these proteins are used for training and 383 of 14 them for test. Overall, classification system achieves 81.2% fold recognition accuracy on 15 the test set, where most of the proteins have less than 25% sequence identity with the ones 16 used during the training. To portray the capabilities of GAL network among the other 17 methods, comparisons between a few approaches have also been made, and GAL's 18 accuracy is found to be higher than those of the existing methods for protein fold 19 classification. 20