Prediction of Seven Protein Structural Classes by Fusing Multi-Feature Information Including Protein Evolutionary Conservation Information

Determination of protein structural class is a quite meaningful topic in protein science, because a priori knowledge of a protein structural class can provide useful information about its overall structure. The results of most previous studies used high homologous dataset with four structural classes should not be perceived as reliable, because the sequence homology has very significant impact on the prediction accuracy. Using a rigorous dataset with only less than 20% sequence identity to each other, this paper developed a novel pseudo amino acid composition method (PseAA) approach by incorporating protein evolutionary conservation information, amino acid physicochemical properties and statistical information to predict seven structural classes. Comparing with another PseAA method, the overall accuracy of our multi-feature information fusion method is 4.6% higher than that of the method of autocorrelation function of amino acid RICJ880103 in jackknife test. The results indicate that multi-feature information fusion including evolutionary information is effective and robust for the prediction of protein structural class with low sequence identity dataset, and can be effectively complemented with functional domain composition method.