Prediction of functional class of proteins and peptides irrespective of sequence homology by support vector machines

Abstract

Various computational methods have been used for the prediction of protein and peptide function based on their sequences. A particular challenge is to derive functional properties from sequences that show low or no homology to proteins of known function. Recently, a machine learning method, support vector machines (SVM), have been explored for predicting functional class of proteins and peptides from amino acid sequence derived properties independent of sequence similarity, which have shown promising potential for a wide spectrum of protein and peptide classes including some of the low- and non-homologous proteins. This method can thus be explored as a potential tool to complement alignment-based, clustering-based, and structure-based methods for predicting protein function. This article reviews the strategies, current progresses, and underlying difficulties in using SVM for predicting the functional class of proteins. The relevant software and web-servers are described. The reported prediction performances in the application of these methods are also presented.

Keywords: machine learning method; peptide function; protein family; protein function; protein function prediction; support vector machines.

Figure 1.

Schematic diagram illustrating the process of the training and prediction of the functional class of proteins and peptides by using support vector machine (SVM) method. A,B: feature vectors of proteins belong to a functional class; E,F: feature vectors of proteins not belong to a functional class. Sequence-derived feature h_j, v_j, p_j, … represents such structural and physicochemical properties as hydrophobicity, polarizability, and volume; or such properties as domain information, subcellular localization, and post-translational (PT) modification profiles etc.

Figure 2.

Support vector machines. (a) Definition of hyper-plane and margin. The circular dots and square dots represent samples of class −1 and class +1, respectively. (b) The available hyper-planes H, H’, H”, …, corresponding to a set of training data. (c) Unique optimal separating hyper-plane of a set of training data. (d) Basic idea of support vector machines: Projection of the training data nonlinearly into a higher-dimensional feature space via φ, and subsequent construction of a separating hyper-plane with maximum margin in that space.