Predicting mouse transmembrane protein types based on the increment of diversity combined with the support vector machine

Transmembrane proteins play important roles in biology. They are especially important in signal reception, molecular pumping and energy transduction. Their medical importance is also growing rapidly after the completion of many large genome projects. In this paper, a mouse transmembrane protein types database was constructed. Based on the concept of representing protein samples in terms of their amino acid compositions and pseudo-amino acid compositions, the increment of diversity (ID) and the support vector machine(SVM) are introduced to predict the mouse transmembrane protein types. On this basis, a new algorithm of the increment of diversity combined with the support vector machine (ID_SVM) is proposed to predict the mouse transmembrane protein types. The overall prediction accuracy in jackknife test is 77.37%. This result is better than the results of the increment of diversity and the support vector machine.