Simple alignment-free methods for protein classification: a case study from G-protein-coupled receptors

Computational methods of predicting protein functions rely on detecting similarities among proteins. However, sufficient sequence information is not always available for some protein families. For example, proteins of interest may be new members of a divergent protein family. The performance of protein classification methods could vary in such challenging situations. Using the G-protein-coupled receptor superfamily as an example, we investigated the performance of several protein classifiers. Alignment-free classifiers based on support vector machines using simple amino acid compositions were effective in remote-similarity detection even from short fragmented sequences. Although it is computationally expensive, a support vector machine classifier using local pairwise alignment scores showed very good balanced performance. More commonly used profile hidden Markov models were generally highly specific and well suited to classifying well-established protein family members. It is suggested that different types of protein classifiers should be applied to gain the optimal mining power.     

Prediction of protein subcellular multi-localization based on the general form of Chou's pseudo amino acid composition

Many proteins bear multi-locational characteristics, and this phenomenon is closely related to biological function. However, most of the existing methods can only deal with single-location proteins. Therefore, an automatic and reliable ensemble classifier for protein subcellular multi-localization is needed. We propose a new ensemble classifier combining the KNN (K-nearest neighbour) and SVM (support vector machine) algorithms to predict the subcellular localization of eukaryotic, Gram-negative bacterial and viral proteins based on the general form of Chou's pseudo amino acid composition, i.e., GO (gene ontology) annotations, dipeptide composition and AmPseAAC (Amphiphilic pseudo amino acid composition). This ensemble classifier was developed by fusing many basic individual classifiers through a voting system. The overall prediction accuracies obtained by the KNN-SVM ensemble classifier are 95.22, 93.47 and 80.72% for the eukaryotic, Gram-negative bacterial and viral proteins, respectively. Our prediction accuracies are significantly higher than those by previous methods and reveal that our strategy better predicts subcellular locations of multi-location proteins.     

PSP_MCSVM: brainstorming consensus prediction of protein secondary structures using two-stage multiclass support vector machines

Secondary structure prediction is a crucial task for understanding the variety of protein structures and performed biological functions. Prediction of secondary structures for new proteins using their amino acid sequences is of fundamental importance in bioinformatics. We propose a novel technique to predict protein secondary structures based on position-specific scoring matrices (PSSMs) and physico-chemical properties of amino acids. It is a two stage approach involving multiclass support vector machines (SVMs) as classifiers for three different structural conformations, viz., helix, sheet and coil. In the first stage, PSSMs obtained from PSI-BLAST and five specially selected physicochemical properties of amino acids are fed into SVMs as features for sequence-to-structure prediction. Confidence values for forming helix, sheet and coil that are obtained from the first stage SVM are then used in the second stage SVM for performing structure-to-structure prediction. The two-stage cascaded classifiers (PSP_MCSVM) are trained with proteins from RS126 dataset. The classifiers are finally tested on target proteins of critical assessment of protein structure prediction experiment-9 (CASP9). PSP_MCSVM with brainstorming consensus procedure performs better than the prediction servers like Predator, DSC, SIMPA96, for randomly selected proteins from CASP9 targets. The overall performance is found to be comparable with the current state-of-the art. PSP_MCSVM source code, train-test datasets and supplementary files are available freely in public domain at: and     

Machine learning approaches for prediction of linear B-cell epitopes on proteins

Identification and characterization of antigenic determinants on proteins has received considerable attention utilizing both, experimental as well as computational methods. For computational routines mostly structural as well as physicochemical parameters have been utilized for predicting the antigenic propensity of protein sites. However, the performance of computational routines has been low when compared to experimental alternatives. Here we describe the construction of machine learning based classifiers to enhance the prediction quality for identifying linear B-cell epitopes on proteins. Our approach combines several parameters previously associated with antigenicity, and includes novel parameters based on frequencies of amino acids and amino acid neighborhood propensities. We utilized machine learning algorithms for deriving antigenicity classification functions assigning antigenic propensities to each amino acid of a given protein sequence. We compared the prediction quality of the novel classifiers with respect to established routines for epitope scoring, and tested prediction accuracy on experimental data available for HIV proteins. The major finding is that machine learning classifiers clearly outperform the reference classification systems on the HIV epitope validation set.     

A critical evaluation of network and pathway-based classifiers for outcome prediction in breast cancer

Recently, several classifiers that combine primary tumor data, like gene expression data, and secondary data sources, such as protein-protein interaction networks, have been proposed for predicting outcome in breast cancer. In these approaches, new composite features are typically constructed by aggregating the expression levels of several genes. The secondary data sources are employed to guide this aggregation. Although many studies claim that these approaches improve classification performance over single genes classifiers, the gain in performance is difficult to assess. This stems mainly from the fact that different breast cancer data sets and validation procedures are employed to assess the performance. Here we address these issues by employing a large cohort of six breast cancer data sets as benchmark set and by performing an unbiased evaluation of the classification accuracies of the different approaches. Contrary to previous claims, we find that composite feature classifiers do not outperform simple single genes classifiers. We investigate the effect of (1) the number of selected features; (2) the specific gene set from which features are selected; (3) the size of the training set and (4) the heterogeneity of the data set on the performance of composite feature and single genes classifiers. Strikingly, we find that randomization of secondary data sources, which destroys all biological information in these sources, does not result in a deterioration in performance of composite feature classifiers. Finally, we show that when a proper correction for gene set size is performed, the stability of single genes sets is similar to the stability of composite feature sets. Based on these results there is currently no reason to prefer prognostic classifiers based on composite features over single genes classifiers for predicting outcome in breast cancer.     

Rapid identification of protein-protein interfaces for the construction of a complex model based on multiple unassigned signals by using time-sharing NMR measurements

Protein-protein interactions are necessary for various cellular processes, and therefore, information related to protein-protein interactions and structural information of complexes is invaluable. To identify protein-protein interfaces using NMR, resonance assignments are generally necessary to analyze the data; however, they are time consuming to collect, especially for large proteins. In this paper, we present a rapid, effective, and unbiased approach for the identification of a protein-protein interface without resonance assignments. This approach requires only a single set of 2D titration experiments of a single protein sample, labeled with a unique combination of an (15)N-labeled amino acid and several amino acids (13)C-labeled on specific atoms. To rapidly obtain high resolution data, we applied a new pulse sequence for time-shared NMR measurements that allowed simultaneous detection of a ω(1)-TROSY-type backbone (1)H-(15)N and aromatic (1)H-(13)C shift correlations together with single quantum methyl (1)H-(13)C shift correlations. We developed a structure-based computational approach, that uses our experimental data to search the protein surfaces in an unbiased manner to identify the residues involved in the protein-protein interface. Finally, we demonstrated that the obtained information of the molecular interface could be directly leveraged to support protein-protein docking studies. Such rapid construction of a complex model provides valuable information and enables more efficient biochemical characterization of a protein-protein complex, for instance, as the first step in structure-guided drug development.     

TSSub: eukaryotic protein subcellular localization by extracting features from profiles

This paper introduces a new subcellular localization system (TSSub) for eukaryotic proteins. This system extracts features from both profiles and amino acid sequences. Four different features are extracted from profiles by four probabilistic neural network (PNN) classifiers, respectively (the amino acid composition from whole profiles; the amino acid composition from the N-terminus of profiles; the dipeptide composition from whole profiles and the amino acid composition from fragments of profiles). In addition, a support vector machine (SVM) classifier is added to implement the residue-couple feature extracted from amino acid sequences. The results from the five classifiers are fused by an additional SVM classifier. The overall accuracies of this TSSub reach 93.0 and 77.4% on Reinhardt and Hubbard's eukaryotic protein dataset and Huang and Li's eukaryotic protein dataset, respectively. The comparison with existing methods results shows TSSub provides better prediction performance than existing methods. AVAILABILITY: The web server is available from http://166.111.24.5/webtools/TSSub/index.html.     

Fragment library-based representation system for protein conformation

The representation system for protein conformation has a crucial effect on the speed of various protein-related simulations, including ab initio protein structure prediction and protein-protein docking simulation. Usually, the finer a representation system, the longer is the computational time required to employ the representation system in simulations. On the other hand, very coarse lattice systems cannot be directly applied to the simulation problems with real proteins. We report a new, fragment library-based protein conformation representation system, prepared by clustering amino acid conformations from 154 proteins. This system was composed of 64 most representative fragments per each amino acid, and based on the unified residue approach in which two spheres per amino acid were used. It could represent the conformation of the 82 proteins in an independent test set with the mean and standard deviation RMSD of 1.01 and 0.09 A, respectively, based on the position of alpha carbons and the centers of mass of sidechains.     

A photocrosslinking assay for reporting protein interactions in polyketide and fatty acid synthases

Understanding protein-protein interactions that occur between ACP and KS domains of polyketide synthases and fatty acid synthases is critical to improving the scope and efficiency of combinatorial biosynthesis efforts aimed at producing non-natural polyketides. Here, we report a facile strategy for rapidly reporting such ACP-KS interactions based on the incorporation of an amino acid with photocrosslinking functionality. Crucially, this photocrosslinking strategy can be applied to any polyketide or fatty acid synthase regardless of substrate specificity, and can be adapted to a high-throughput format for directed evolution studies.     

Random projection ensemble classification with high-dimensional time series

Multivariate time-series (MTS) data are prevalent in diverse domains and often high dimensional. We propose new random projection ensemble classifiers with high-dimensional MTS. The method first applies dimension reduction in the time domain via randomly projecting the time-series variables into some low-dimensional space, followed by measuring the disparity via some novel base classifier between the data and the candidate generating processes in the projected space. Our contributions are twofold: (i) We derive optimal weighted majority voting schemes for pooling information from the base classifiers for multiclass classification and (ii) we introduce new base frequency-domain classifiers based on Whittle likelihood (WL), Kullback-Leibler (KL) divergence, eigen-distance (ED), and Chernoff (CH) divergence. Both simulations for binary and multiclass problems, and an Electroencephalogram (EEG) application demonstrate the efficacy of the proposed methods in constructing accurate classifiers with high-dimensional MTS.     

Prediction of turn types in protein structure by machine-learning classifiers

We present machine learning approaches for turn prediction from the amino acid sequence. Different turn classes and types were considered based on a novel turn classification scheme. We trained an unsupervised (self-organizing map) and two kernel-based classifiers, namely the support vector machine and a probabilistic neural network. Turn versus non-turn classification was carried out for turn families containing intramolecular hydrogen bonds and three to six residues. Support vector machine classifiers yielded a Matthews correlation coefficient (mcc) of approximately 0.6 and a prediction accuracy of 80%. Probabilistic neural networks were developed for beta-turn type prediction. The method was able to distinguish between five types of beta-turns yielding mcc > 0.5 and at least 80% overall accuracy. We conclude that the proposed new turn classification is distinct and well-defined, and machine learning classifiers are suited for sequence-based turn prediction. Their potential for sequence-based prediction of turn structures is discussed.     

Identification of a PA-Binding Peptide with Inhibitory Activity against Influenza A and B Virus Replication

There is an urgent need for new drugs against influenza type A and B viruses due to incomplete protection by vaccines and the emergence of resistance to current antivirals. The influenza virus polymerase complex, consisting of the PB1, PB2 and PA subunits, represents a promising target for the development of new drugs. We have previously demonstrated the feasibility of targeting the protein-protein interaction domain between the PB1 and PA subunits of the polymerase complex of influenza A virus using a small peptide derived from the PA-binding domain of PB1. However, this influenza A virus-derived peptide did not affect influenza B virus polymerase activity. Here we report that the PA-binding domain of the polymerase subunit PB1 of influenza A and B viruses is highly conserved and that mutual amino acid exchange shows that they cannot be functionally exchanged with each other. Based on phylogenetic analysis and a novel biochemical ELISA-based screening approach, we were able to identify an influenza A-derived peptide with a single influenza B-specific amino acid substitution which efficiently binds to PA of both virus types. This dual-binding peptide blocked the viral polymerase activity and growth of both virus types. Our findings provide proof of principle that protein-protein interaction inhibitors can be generated against influenza A and B viruses. Furthermore, this dual-binding peptide, combined with our novel screening method, is a promising platform to identify new antiviral lead compounds.     

Taking advantage of local structure descriptors to analyze interresidue contacts in protein structures and protein complexes

Interresidue protein contacts in proteins structures and at protein-protein interface are classically described by the amino acid types of interacting residues and the local structural context of the contact, if any, is described using secondary structures. In this study, we present an alternate analysis of interresidue contact using local structures defined by the structural alphabet introduced by Camproux et al. This structural alphabet allows to describe a 3D structure as a sequence of prototype fragments called structural letters, of 27 different types. Each residue can then be assigned to a particular local structure, even in loop regions. The analysis of interresidue contacts within protein structures defined using Vorono? tessellations reveals that pairwise contact specificity is greater in terms of structural letters than amino acids. Using a simple heuristic based on specificity score comparison, we find that 74% of the long-range contacts within protein structures are better described using structural letters than amino acid types. The investigation is extended to a set of protein-protein complexes, showing that the similar global rules apply as for intraprotein contacts, with 64% of the interprotein contacts best described by local structures. We then present an evaluation of pairing functions integrating structural letters to decoy scoring and show that some complexes could benefit from the use of structural letter-based pairing functions.     

Efficient Kernelized prototype based classification

Prototype based classifiers are effective algorithms in modeling classification problems and have been applied in multiple domains. While many supervised learning algorithms have been successfully extended to kernels to improve the discrimination power by means of the kernel concept, prototype based classifiers are typically still used with Euclidean distance measures. Kernelized variants of prototype based classifiers are currently too complex to be applied for larger data sets. Here we propose an extension of Kernelized Generalized Learning Vector Quantization (KGLVQ) employing a sparsity and approximation technique to reduce the learning complexity. We provide generalization error bounds and experimental results on real world data, showing that the extended approach is comparable to SVM on different public data.