Biological Features for Sequence-Based Prediction of Protein Stability Changes upon Amino Acid Substitutions

Abstract

Protein destabilization is a common mechanism by which amino acid substitutions cause human diseases. In this study, a new machine learning method has been developed for sequence-based prediction of protein stability changes upon single amino acid substitutions. Support vector machines were trained with data from experimental studies on the free energy change of protein stability upon mutations. To construct accurate classifiers, twenty biological features were examined for input vector encoding. It was shown that classifier performance varied significantly by the use of different features. The most accurate classifier was constructed using a combination of several biological features. This classifier achieved an overall accuracy of 82.24% with 75.24% sensitivity and 85.36% specificity. Predictive results at this level of accuracy may be used in human genetic studies to distinguish between deleterious and tolerant alterations in disease candidate genes.