An empirical study on the matrix-based protein representations and their combination with sequence-based approaches

Abstract

Many domains have a stake in the development of reliable systems for automatic protein classification. Of particular interest in recent studies of automatic protein classification is the exploration of new methods for extracting features from a protein that enhance classification for specific problems. These methods have proven very useful in one or two domains, but they have failed to generalize well across several domains (i.e. classification problems). In this paper, we evaluate several feature extraction approaches for representing proteins with the aim of sequence-based protein classification. Several protein representations are evaluated, those starting from: the position specific scoring matrix (PSSM) of the proteins; the amino-acid sequence; a matrix representation of the protein, of dimension (length of the protein) ×20, obtained using the substitution matrices for representing each amino-acid as a vector. A valuable result is that a texture descriptor can be extracted from the PSSM protein representation which improves the performance of standard descriptors based on the PSSM representation. Experimentally, we develop our systems by comparing several protein descriptors on nine different datasets. Each descriptor is used to train a support vector machine (SVM) or an ensemble of SVM. Although different stand-alone descriptors work well on some datasets (but not on others), we have discovered that fusion among classifiers trained using different descriptors obtains a good performance across all the tested datasets. Matlab code/Datasets used in the proposed paper are available at http://www.bias.csr.unibo.it\nanni\PSSM.rar.

Gram-positive and Gram-negative protein subcellular localization by incorporating evolutionary-based descriptors into Chou׳s general PseAAC

“…To tackle this problem, a wide range of classification techniques have been implemented and used Shen, 2007a, 2006a;Chou and Shen, 2008;Wan et al, 2013;Chou and Cai, 2002;Yu et al, 2013;Nanni et al, 2013aNanni et al, , 2013bShen and Chou, 2007;Huang and Yuan, 2013). Among these classifiers, Support Vector Machine (SVM) (Wan et al, 2013;Yu et al, 2013;Pierleoni et al, 2011;Du and Yu, 2013;Matsuda et al, 2005) or K-Nearest Neighbor (KNN) based classifiers Shen and Chou, 2010a;Chen et al, 2013aChen et al, , 2012 have attained the most promising results.…”

“…The main reason is that previous studies failed to capture local discriminatory information embedded in PSSM properly. They have mainly tried to extract this local information using the protein sequence as a single block which has failed to achieve this goal (Nanni et al, 2013b(Nanni et al, , 2013aDehzangi et al, 2014a).…”

A Segmentation-Based Method to Extract Structural and Evolutionary Features for Protein Fold Recognition

“…During the last decade, the substitution score (extracted from the PSSM) and predicted secondary structure using PSIBLAST and SPINE-X (or PSIPRED before that) have been widely used in protein science (e.g., protein fold recognition, protein function prediction, protein structure prediction, protein subcellular localization) and extracted features from these sources attained promising results [23], [28], [30], [38], [42], [48]. As it is highlighted in [49], the most sensitive methods for fold recognition use sequence profiles to represent both the query and the data base proteins.…”

“…Compared to the methods adopted to extract global discriminatory information, a wider range of methods were used to extract local discriminatory information for the PFR [42] such as, pseudo amino acid composition [14], [15], [22], [23], [43], cross covariance [28], auto covariance [28], [42], bi-gram [29], [39], and tri-gram [3]. Despite the significant local discriminatory information provided using these approaches, most of these methods produce large number of features as well as large amount of redundant features [14], [29] which makes them computationally expensive for large protein data banks (e.g., cross covariance and tri-gram [3], [28]).…”

MitProt-Pred: Predicting mitochondrial proteins of Plasmodium falciparum parasite using diverse physiochemical properties and ensemble classification

“…Such techniques use sequence and biological information simultaneously. Several works are already proposed for sequence based classification using ensemble [4,[6][7][8]. In addition, there exist various techniques including Target P, Signal P3.0, WoLF PSORT, TargetLoc, MitoProt II, MITOPRED, MitPred, and Mito-GSAAC, which are all organelle specific methods [4,[9][10][11][12][13][14][15].…”