Prediction of Multi-Type Membrane Proteins in Human by an Integrated Approach

Abstract

Membrane proteins were found to be involved in various cellular processes performing various important functions, which are mainly associated to their types. However, it is very time-consuming and expensive for traditional biophysical methods to identify membrane protein types. Although some computational tools predicting membrane protein types have been developed, most of them can only recognize one kind of type. Therefore, they are not as effective as one membrane protein can have several types at the same time. To our knowledge, few methods handling multiple types of membrane proteins were reported. In this study, we proposed an integrated approach to predict multiple types of membrane proteins by employing sequence homology and protein-protein interaction network. As a result, the prediction accuracies reached 87.65%, 81.39% and 70.79%, respectively, by the leave-one-out test on three datasets. It outperformed the nearest neighbor algorithm adopting pseudo amino acid composition. The method is anticipated to be an alternative tool for identifying membrane protein types. New metrics for evaluating performances of methods dealing with multi-label problems were also presented. The program of the method is available upon request.

Classification of membrane protein types using Voting Feature Interval in combination with Chou׳s Pseudo Amino Acid Composition

“…In case of dataset S1, NNA based method has obtained 70.41% accuracy, RWC has yielded 81.34% accuracy whereas integrated method has achieved 87.65% accuracy. In contrast, our proposed method has obtained 93.9% accuracy which is 6.25% higher than the highest result of Huang et al, method. Similarly in case of dataset S2, the Huang et al, methods have achieved 61.70%, 71.40%, and 71.79% accuracy on NNA based method, RWC, and integrated method, respectively (Huang et al, 2014). On the other hand, our proposed method has yielded 89.33% accuracy which is 17.54% higher than existing methods in the literature.…”

“…On the other hand, our proposed method has yielded 89.33% accuracy which is 17.54% higher than existing methods in the literature. In case of dataset S3, the Huang et al, methods have yielded 56.33% accuracy on NNA based method, 56.82% accuracy on RWC method, and 70.79% accuracy on integrated method (Huang et al, 2014). Whereas, our proposed method has obtained 86.9% accuracy, which is 16.11% higher than the highest accuracy of existing methods in the literature.…”

“…Huang et al, have applied three techniques namely nearest neighbor algorithm (NNA), real weighted combination (RWC) and integrated method in all the three datasets (Huang et al, 2014). In case of dataset S1, NNA based method has obtained 70.41% accuracy, RWC has yielded 81.34% accuracy whereas integrated method has achieved 87.65% accuracy.…”

“…So knowledge of membrane protein's types provides great help in determining its functions. It is reported that membrane proteins have generally six types (Huang et al, 2014) : (1) GPI-anchor; (2) lipid-anchor; (3) multi-pass; (4) peripheral; (5) single-pass type I; (6) single-pass type II membrane proteins. To predict the types of membrane protein numerous bioinformatics methods were developed.…”

An efficient approach for the prediction of ion channels and their subfamilies

“…Ion channels play an important target for antiepileptic drug design, antihypertensive and antipsychotics disorder such as schizophrenia [3,4]. Currently, there are various computational intelligence techniques based approaches that have been proposed to predict membrane proteins [5][6][7][8][9][10][11] .Similarly various computational intelligence techniques based approaches have been proposed to predict nuclear receptor [12][13][14][15][16][17] and G-protein coupled receptor [18][19][20][21][22].…”