An evaluation of protein secondary structure prediction algorithms

Abstract

Publisher Summary Algorithms for secondary structure prediction are based upon diverse theoretical approaches. Over the time, several algorithms have been developed to predict the secondary structure of proteins based on very distinct theoretical approaches. Comparative analysis of the performance of various algorithms can be deceptive if factors, such as the selection of proteins for the testing set and the choice of the scoring Index are not carried out properly. The chapter provides an updated evaluation of several predictive methods with a testing set size that permits to obtain more accurate statistics, which in turn can possibly measure the usefulness of the information gathered by those methods and identify trends that characterize the behavior of individual algorithms. Further, the chapter presents a uniform testing of these methods, vis-a-vis the size of the datasets, the measure of accuracy and proper cross-validation procedures. From a large list of available algorithms for secondary structure prediction, nine are selected to represent the main classes. These are chosen mainly because they are the most commonly cited in the literature and by the fact that they permit a relatively safe implementation through of a computer program. The selected methods are summarized in the chapter. The present analysis might give rise to a somewhat pessimistic view of the effectiveness of protein secondary structure prediction algorithms. In fact, with the increasing number of proteins with known three-dimensional structure, constant re-evaluation of the performance must take place to ascertain the validity of the methods. It is noted that the methods are analyzed consistently using the 148 proteins selected in this study. Moreover, the situation is even worse for the Mathews correlation coefficient, which indicates that the predictions are poorly correlated with the actual structure.