Feature extraction of protein folds based on secondary structure transformation

Abstract

We propose a classification method of protein folds from the structural features of clusters that are extracted based on the cost of structural transformation from one protein to another. In the method, the feature to be extracted is defined as a pair of a delegate structure and a cluster region. The delegate structure is generated by estimating the center of structural balance among proteins in one cluster from the viewpoint of transformation costs. The cluster region shows how far the cluster spread. The validity of the proposed feature extraction method is confirmed through the application to the convenient classification of unclassified proteins.