SVM-based method for protein structural class prediction using secondary structural content and structural information of amino acids

The knowledge collated from the known protein structures has revealed that the proteins are usually folded into the four structural classes: all-α, all-β, α/β and α + β. A number of methods have been proposed to predict the protein's structural class from its primary structure; however, it has been observed that these methods fail or perform poorly in the cases of distantly related sequences. In this paper, we propose a new method for protein structural class prediction using low homology (twilight-zone) protein sequences dataset. Since protein structural class prediction is a typical classification problem, we have developed a Support Vector Machine (SVM)-based method for protein structural class prediction that uses features derived from the predicted secondary structure and predicted burial information of amino acid residues. The examination of different individual as well as feature combinations revealed that the combination of secondary structural content, secondary structural and solvent accessibility state frequencies of amino acids gave rise to the best leave-one-out cross-validation accuracy of ~81% which is comparable to the best accuracy reported in the literature so far.     

Accuracy of sequence alignment and fold assessment using reduced amino acid alphabets

Reduced or simplified amino acid alphabets group the 20 naturally occurring amino acids into a smaller number of representative protein residues. To date, several reduced amino acid alphabets have been proposed, which have been derived and optimized by a variety of methods. The resulting reduced amino acid alphabets have been applied to pattern recognition, generation of consensus sequences from multiple alignments, protein folding, and protein structure prediction. In this work, amino acid substitution matrices and statistical potentials were derived based on several reduced amino acid alphabets and their performance assessed in a large benchmark for the tasks of sequence alignment and fold assessment of protein structure models, using as a reference frame the standard alphabet of 20 amino acids. The results showed that a large reduction in the total number of residue types does not necessarily translate into a significant loss of discriminative power for sequence alignment and fold assessment. Therefore, some definitions of a few residue types are able to encode most of the relevant sequence/structure information that is present in the 20 standard amino acids. Based on these results, we suggest that the use of reduced amino acid alphabets may allow to increasing the accuracy of current substitution matrices and statistical potentials for the prediction of protein structure of remote homologs.     

The recognition of multi-class protein folds by adding average chemical shifts of secondary structure elements

The recognition of protein folds is an important step in the prediction of protein structure and function. Recently, an increasing number of researchers have sought to improve the methods for protein fold recognition. Following the construction of a dataset consisting of 27 protein fold classes by Ding and Dubchak in 2001, prediction algorithms, parameters and the construction of new datasets have improved for the prediction of protein folds. In this study, we reorganized a dataset consisting of 76-fold classes constructed by Liu et al. and used the values of the increment of diversity, average chemical shifts of secondary structure elements and secondary structure motifs as feature parameters in the recognition of multi-class protein folds. With the combined feature vector as the input parameter for the Random Forests algorithm and ensemble classification strategy, we propose a novel method to identify the 76 protein fold classes. The overall accuracy of the test dataset using an independent test was 66.69%; when the training and test sets were combined, with 5-fold cross-validation, the overall accuracy was 73.43%. This method was further used to predict the test dataset and the corresponding structural classification of the first 27-protein fold class dataset, resulting in overall accuracies of 79.66% and 93.40%, respectively. Moreover, when the training set and test sets were combined, the accuracy using 5-fold cross-validation was 81.21%. Additionally, this approach resulted in improved prediction results using the 27-protein fold class dataset constructed by Ding and Dubchak.     

Prediction of the parallel/antiparallel orientation of beta-strands using amino acid pairing preferences and support vector machines

In principle, structural information of protein sequences with no detectable homology to a protein of known structure could be obtained by predicting the arrangement of their secondary structural elements. Although some ab initio methods for protein structure prediction have been reported, the long-range interactions required to accurately predict tertiary structures of β-sheet containing proteins are still difficult to simulate. To remedy this problem and facilitate de novo prediction of β-sheet containing protein structures, we developed a support vector machine (SVM) approach that classified parallel and antiparallel orientation of β-strands by using the information of interstrand amino acid pairing preferences. Based on a second-order statistics on the relative frequencies of each possible interstrand amino acid pair, we defined an average amino acid pairing encoding matrix (APEM) for encoding β-strands as input in the prediction model. As a result, a prediction accuracy of 86.89% and a Matthew's correlation coefficient value of 0.71 have been achieved through 7-fold cross-validation on a non-redundant protein dataset from PISCES. Although several issues still remain to be studied, the method presented here to some extent could indicate the important contribution of the amino acid pairs to the β-strand orientation, and provide a possible way to further be combined with other algorithms making a full ‘identification’ of β-strands.     

Environment-specific amino acid substitution tables: tertiary templates and prediction of protein folds.

The local environment of an amino acid in a folded protein determines the acceptability of mutations at that position. In order to characterize and quantify these structural constraints, we have made a comparative analysis of families of homologous proteins. Residues in each structure are classified according to amino acid type, secondary structure, accessibility of the side chain, and existence of hydrogen bonds from the side chains. Analysis of the pattern of observed substitutions as a function of local environment shows that there are distinct patterns, especially for buried polar residues. The substitution data tables are available on diskette with Protein Science. Given the fold of a protein, one is able to predict sequences compatible with the fold (profiles or templates) and potentially to discriminate between a correctly folded and misfolded protein. Conversely, analysis of residue variation across a family of aligned sequences in terms of substitution profiles can allow prediction of secondary structure or tertiary environment.     

Modeling amino acid replacement.

The estimation of amino acid replacement frequencies during molecular evolution is crucial for many applications in sequence analysis. Score matrices for database search programs or phylogenetic analysis rely on such models of protein evolution. Pioneering work was done by Dayhoff et al. (1978) who formulated a Markov model of evolution and derived the famous PAM score matrices. Her estimation procedure for amino acid exchange frequencies is restricted to pairs of proteins that have a constant and small degree of divergence. Here we present an improved estimator, called the resolvent method, that is not subject to these limitations. This extension of Dayhoff's approach enables us to estimate an amino acid substitution model from alignments of varying degree of divergence. Extensive simulations show the capability of the new estimator to recover accurately the exchange frequencies among amino acids. Based on the SYSTERS database of aligned protein families (Krause and Vingron, 1998) we recompute a series of score matrices.     

Predicting protein fold types by the general form of Chou's pseudo amino acid composition: approached from optimal feature extractions

Identification on protein folding types is always based on the 27-class folds dataset, which was provided by Ding & Dubchak in 2001. But with the avalanche of protein sequences, fold data is also expanding, so it will be the inevitable trend to improve the existing dataset and expand more folding types. In this paper, we construct a multi-class protein fold dataset, which contains 3,457 protein chains with sequence identity below 35% and could be classified into 76 fold types. It was 4 times larger than Ding & Dubchak's dataset. Furthermore, our work proposes a novel approach of support vector machine based on optimal features. By combining motif frequency, low-frequency power spectral density, amino acid composition, the predicted secondary structure and the values of auto-correlation function as feature parameters set, the method adopts criterion of the maximum correlation and the minimum redundancy to filter these features and obtain a 95-dimensions optimal feature subset. Based on the ensemble classification strategy, with 95-dimensions optimal feature as input parameters of support vector machine, we identify the 76-class protein folds and overall accuracy measures up to 44.92% by independent test. In addition, this method has been further used to identify upgraded 27-class protein folds, overall accuracy achieves 66.56%. At last, we also test our method on Ding & Dubchak's 27-class folds dataset and obtained better identification results than most of the previous reported results.     

Correlated substitution analysis and the prediction of amino acid structural contacts

It has long been suspected that analysis of correlated amino acid substitutions should uncover pairs or clusters of sites that are spatially proximal in mature protein structures. Accordingly, methods based on different mathematical principles such as information theory, correlation coefficients and maximum likelihood have been developed to identify co-evolving amino acids from multiple sequence alignments. Sets of pairs of sites whose behaviour is identified by these methods as correlated are often significantly enriched in pairs of spatially proximal residues. However, relatively high levels of false-positive predictions typically render such methods, in isolation, of little use in the ab initio prediction of protein structure. Misleading signal (or problems with the estimation of significance levels) can be caused by phylogenetic correlations between homologous sequences and from correlation due to factors other than spatial proximity (for example, correlation of sites which are not spatially close but which are involved in common functional properties of the protein). In recent years, several workers have suggested that information from correlated substitutions should be combined with other sources of information (secondary structure, solvent accessibility, evolutionary rates) in an attempt to reduce the proportion of false-positive predictions. We review methods for the detection of correlated amino acid substitutions, compare their relative performance in contact prediction and predict future directions in the field.     

基于氨基酸组成分布的蛋白质同源寡聚体分类研究

基于一种新的特征提取方法——氨基酸组成分布,使用支持向量机作为成员分类器,采用“一对一”的多类分类策略,从蛋白质一级序列对四类同源寡聚体进行分类研究。结果表明,在10-CV检验下,基于氨基酸组成分布,其总分类精度和精度指数分别达到了86.22%和67.12%,比基于氨基酸组成成分的传统特征提取方法分别提高了5.74和10.03个百分点,比二肽组成成分特征提取方法分别提高了3.12和5.63个百分点,说明氨基酸组成分布对于蛋白质同源寡聚体分类是一种非常有效的特征提取方法;将氨基酸组成分布和蛋白质序列长度特征组合,其总分类精度和精度指数分别达到了86.35%和67.23%,说明蛋白质序列长度特征含有一定的空间结构信息。     

Prediction of protein mutant stability using classification and regression tool

Prediction of protein stability upon amino acid substitutions is an important problem in molecular biology and the solving of which would help for designing stable mutants. In this work, we have analyzed the stability of protein mutants using two different datasets of 1396 and 2204 mutants obtained from ProTherm database, respectively for free energy change due to thermal (DeltaDeltaG) and denaturant denaturations (DeltaDeltaG(H(2)O)). We have used a set of 48 physical, chemical energetic and conformational properties of amino acid residues and computed the difference of amino acid properties for each mutant in both sets of data. These differences in amino acid properties have been related to protein stability (DeltaDeltaG and DeltaDeltaG(H(2)O)) and are used to train with classification and regression tool for predicting the stability of protein mutants. Further, we have tested the method with 4 fold, 5 fold and 10 fold cross validation procedures. We found that the physical properties, shape and flexibility are important determinants of protein stability. The classification of mutants based on secondary structure (helix, strand, turn and coil) and solvent accessibility (buried, partially buried, partially exposed and exposed) distinguished the stabilizing/destabilizing mutants at an average accuracy of 81% and 80%, respectively for DeltaDeltaG and DeltaDeltaG(H(2)O). The correlation between the experimental and predicted stability change is 0.61 for DeltaDeltaG and 0.44 for DeltaDeltaG(H(2)O). Further, the free energy change due to the replacement of amino acid residue has been predicted within an average error of 1.08 kcal/mol and 1.37 kcal/mol for thermal and chemical denaturation, respectively. The relative importance of secondary structure and solvent accessibility, and the influence of the dataset on prediction of protein mutant stability have been discussed.     

Prediction of secondary structural content of proteins from their amino acid composition alone. I. New analytic vector decomposition methods

The predictive limits of the amino acid composition for the secondary structural content (percentage of residues in the secondary structural states helix, sheet, and coil) in proteins are assessed quantitatively. For the first time, techniques for prediction of secondary structural content are presented which rely on the amino acid composition as the only information on the query protein. In our first method, the amino acid composition of an unknown protein is represented by the best (in a least square sense) linear combination of the characteristic amino acid compositions of the three secondary structural types computed from a learning set of tertiary structures. The second technique is a generalization of the first one and takes into account also possible compositional couplings between any two sorts of amino acids. Its mathematical formulation results in an eigenvalue/eigenvector problem of the second moment matrix describing the amino acid compositional fluctuations of secondary structural types in various proteins of a learning set. Possible correlations of the principal directions of the eigenspaces with physical properties of the amino acids were also checked. For example, the first two eigenvectors of the helical eigenspace correlate with the size and hydrophobicity of the residue types respectively. As learning and test sets of tertiary structures, we utilized representative, automatically generated subsets of Protein Data Bank (PDB) consisting of non-homologous protein structures at the resolution thresholds ≤1.8Å, ≤2.0Å, ≤2.5Å, and ≤3.0Å. We show that the consideration of compositional couplings improves prediction accuracy, albeit not dramatically. Whereas in the self-consistency test (learning with the protein to be predicted), a clear decrease of prediction accuracy with worsening resolution is observed, the jackknife test (leave the predicted protein out) yielded best results for the largest dataset (≤3.0 Å, almost no difference to the self-consistency test!), i.e., only this set, with more than 400 proteins, is sufficient for stable computation of the parameters in the prediction function of the second method. The average absolute error in predicting the fraction of helix, sheet, and coil from amino acid composition of the query protein are 13.7, 12.6, and 11.4%, respectively with r.m.s. deviations in the range of 8.6 ÷ 11.8% for the 3.0 Å dataset in a jackknife test. The absolute precision of the average absolute errors is in the range of 1 ÷ 3% as measured for other representative subsets of the PDB. Secondary structural content prediction methods found in the literature have been clustered in accordance with their prediction accuracies. To our surprise, much more complex secondary structure prediction methods utilized for the same purpose of secondary structural content prediction achieve prediction accuracies very similar to those of the present analytic techniques, implying that all the information beyond the amino acid composition is, in fact, mainly utilized for positioning the secondary structural state in the sequence but not for determination of the overall number of residues in a secondary structural type. This result implies that higher prediction accuracies cannot be achieved relying solely on the amino acid composition of an unknown query protein as prediction input. Our prediction program SSCP has been made available as a World Wide Web and E-mail service. © 1996 Wiley-Liss, Inc.     

Statistical potential-based amino acid similarity matrices for aligning distantly related protein sequences

Aligning distantly related protein sequences is a long-standing problem in bioinformatics, and a key for successful protein structure prediction. Its importance is increasing recently in the context of structural genomics projects because more and more experimentally solved structures are available as templates for protein structure modeling. Toward this end, recent structure prediction methods employ profile-profile alignments, and various ways of aligning two profiles have been developed. More fundamentally, a better amino acid similarity matrix can improve a profile itself; thereby resulting in more accurate profile-profile alignments. Here we have developed novel amino acid similarity matrices from knowledge-based amino acid contact potentials. Contact potentials are used because the contact propensity to the other amino acids would be one of the most conserved features of each position of a protein structure. The derived amino acid similarity matrices are tested on benchmark alignments at three different levels, namely, the family, the superfamily, and the fold level. Compared to BLOSUM45 and the other existing matrices, the contact potential-based matrices perform comparably in the family level alignments, but clearly outperform in the fold level alignments. The contact potential-based matrices perform even better when suboptimal alignments are considered. Comparing the matrices themselves with each other revealed that the contact potential-based matrices are very different from BLOSUM45 and the other matrices, indicating that they are located in a different basin in the amino acid similarity matrix space.     

A survey of amino acid side-chain interactions in 21 proteins

Based on the atomic co-ordinate data for 21 representative proteins, the frequencies of long-range interactions between side-chain groups and 15 different types of side-chain atoms have been determined. The observed frequencies are compared to the results expected for random association in order to define a scale of relative affinities. Thirty-five residue-atom pairs exhibit frequencies of interaction that differ by at least 50% from the expected values. The amino acids tend to fall into three classes: non-polar, neutral and polar amino acids. The data are regrouped in a different way to determine the average affinity of each amino acid side-chain group for all other types of side-chain groups. Fourteen side-chain pairs have at least 50% fewer interactions than expected, while 21 side-chain pairs have at least 50% more interactions than expected. Unusual patterns of association are discussed and compared with current ideas about the organization of protein structure.     

Conserved key amino acid positions (CKAAPs) derived from the analysis of common substructures in proteins

An all-against-all protein structure comparison using the Combinatorial Extension (CE) algorithm applied to a representative set of PDB structures revealed a gallery of common substructures in proteins (http://cl.sdsc.edu/ce.html). These substructures represent commonly identified folds, domains, or components thereof. Most of the subsequences forming these similar substructures have no significant sequence similarity. We present a method to identify conserved amino acid positions and residue-dependent property clusters within these subsequences starting with structure alignments. Each of the subsequences is aligned to its homologues in SWALL, a nonredundant protein sequence database. The most similar sequences are purged into a common frequency matrix, and weighted homologues of each one of the subsequences are used in scoring for conserved key amino acid positions (CKAAPs). We have set the top 20% of the high-scoring positions in each substructure to be CKAAPs. It is hypothesized that CKAAPs may be responsible for the common folding patterns in either a local or global view of the protein-folding pathway. Where a significant number of structures exist, CKAAPs have also been identified in structure alignments of complete polypeptide chains from the same protein family or superfamily. Evidence to support the presence of CKAAPs comes from other computational approaches and experimental studies of mutation and protein-folding experiments, notably the Paracelsus challenge. Finally, the structural environment of CKAAPs versus non-CKAAPs is examined for solvent accessibility, hydrogen bonding, and secondary structure. The identification of CKAAPs has important implications for protein engineering, fold recognition, modeling, and structure prediction studies and is dependent on the availability of structures and an accurate structure alignment methodology. Proteins 2001;42:148-163.     

Patterns,structures, and amino acid frequencies in structural building blocks,a protein secondary structure classification scheme

To study local structures in proteins, we previously developed an autoassociative artificial neural network (autoANN) and clustering tool to discover intrinsic features of macromolecular structures. The hidden unit activations computed by the trained autoANN are a convenient low-dimensional encoding of the local protein backbone structure. Clustering these activation vectors results in a unique classification of protein local structural features called Structural Building Blocks (SBBs). Here we describe application of this method to a larger database of proteins, verification of the applicability of this method to structure classification, and subsequent analysis of amino acid frequencies and several commonly occurring patterns of SBBs. The SBB classification method has several interesting properties: 1) it identifies the regular secondary structures, α helix and β strand; 2) it consistently identifies other local structure features (e.g., helix caps and strand caps); 3) strong amino acid preferences are revealed at some positions in some SBBs; and 4) distinct patterns of SBBs occur in the “random coil” regions of proteins. Analysis of these patterns identifies interesting structural motifs in the protein backbone structure, indicating that SBBs can be used as “building blocks” in the analysis of protein structure. This type of pattern analysis should increase our understanding of the relationship between protein sequence and local structure, especially in the prediction of protein structures. © 1997 Wiley-Liss, Inc.     

Factors limiting the performance of prediction-based fold recognition methods

In the past few years, a new generation of fold recognition methods has been developed, in which the classical sequence information is combined with information obtained from secondary structure and, sometimes, accessibility predictions. The results are promising, indicating that this approach may compete with potential-based methods (Rost B et al., 1997, J Mol Biol 270:471-480). Here we present a systematic study of the different factors contributing to the performance of these methods, in particular when applied to the problem of fold recognition of remote homologues. Our results indicate that secondary structure and accessibility prediction methods have reached an accuracy level where they are not the major factor limiting the accuracy of fold recognition. The pattern degeneracy problem is confirmed as the major source of error of these methods. On the basis of these results, we study three different options to overcome these limitations: normalization schemes, mapping of the coil state into the different zones of the Ramachandran plot, and post-threading graphical analysis.     

Combining evolutionary and structural information for local protein structure prediction

We study the effects of various factors in representing and combining evolutionary and structural information for local protein structural prediction based on fragment selection. We prepare databases of fragments from a set of non-redundant protein domains. For each fragment, evolutionary information is derived from homologous sequences and represented as estimated effective counts and frequencies of amino acids (evolutionary frequencies) at each position. Position-specific amino acid preferences called structural frequencies are derived from statistical analysis of discrete local structural environments in database structures. Our method for local structure prediction is based on ranking and selecting database fragments that are most similar to a target fragment. Using secondary structure type as a local structural property, we test our method in a number of settings. The major findings are: (1) the COMPASS-type scoring function for fragment similarity comparison gives better prediction accuracy than three other tested scoring functions for profile-profile comparison. We show that the COMPASS-type scoring function can be derived both in the probabilistic framework and in the framework of statistical potentials. (2) Using the evolutionary frequencies of database fragments gives better prediction accuracy than using structural frequencies. (3) Finer definition of local environments, such as including more side-chain solvent accessibility classes and considering the backbone conformations of neighboring residues, gives increasingly better prediction accuracy using structural frequencies. (4) Combining evolutionary and structural frequencies of database fragments, either in a linear fashion or using a pseudocount mixture formula, results in improvement of prediction accuracy. Combination at the log-odds score level is not as effective as combination at the frequency level. This suggests that there might be better ways of combining sequence and structural information than the commonly used linear combination of log-odds scores. Our method of fragment selection and frequency combination gives reasonable results of secondary structure prediction tested on 56 CASP5 targets (average SOV score 0.77), suggesting that it is a valid method for local protein structure prediction. Mixture of predicted structural frequencies and evolutionary frequencies improve the quality of local profile-to-profile alignment by COMPASS.     

Protein thermostability: structure-based difference of amino acid between thermophilic and mesophilic proteins

Structural distributions of each amino acid were compared between 20 pairs of thermophilic and mesophilic proteins to obtain thermostable factors. Five kinds of residual structure states such as fully-exposed, exposed, partially exposed (or partially buried), buried, well-buried states were considered for analyzing the structural patterns of amino acids. The statistical tests revealed that lower frequency in partially exposed state of SER, lower frequency in exposed state and higher frequency in well-buried state of ALA, higher frequency in buried state of GLU, higher frequency in exposed state of ARG, etc. could be critical factors related with protein thermostability.