Predicting DNA- and RNA-binding proteins from sequences with kernel methods

In this paper, support vector machines (SVMs) are applied to predict the nucleic-acid-binding proteins. We constructed two classifiers to differentiate DNA/RNA-binding proteins from non-nucleic-acid-binding proteins by using a conjoint triad feature which extract information directly from amino acids sequence of protein. Both self-consistency and jackknife tests show promising results on the protein datasets in which the sequences identity is less than 25%. In the self-consistency test, the predictive accuracy is 90.37% for DNA-binding proteins and 89.70% for RNA-binding proteins. In the jackknife test, the predictive accuracies are 78.93% and 76.75%, respectively. Comparison results show that our method is very competitive by outperforming other previously published sequence-based prediction methods.     

Prediction of beta-turns with learning machines

The support vector machine approach was introduced to predict the beta-turns in proteins. The overall self-consistency rate by the re-substitution test for the training or learning dataset reached 100%. Both the training dataset and independent testing dataset were taken from Chou [J. Pept. Res. 49 (1997) 120]. The success prediction rates by the jackknife test for the beta-turn subset of 455 tetrapeptides and non-beta-turn subset of 3807 tetrapeptides in the training dataset were 58.1 and 98.4%, respectively. The success rates with the independent dataset test for the beta-turn subset of 110 tetrapeptides and non-beta-turn subset of 30,231 tetrapeptides were 69.1 and 97.3%, respectively. The results obtained from this study support the conclusion that the residue-coupled effect along a tetrapeptide is important for the formation of a beta-turn.     

Support vector machines for prediction of protein subcellular location by incorporating quasi-sequence-order effect.

Support Vector Machine (SVM), which is one class of learning machines, was applied to predict the subcellular location of proteins by incorporating the quasi-sequence-order effect (Chou [2000] Biochem. Biophys. Res. Commun. 278:477-483). In this study, the proteins are classified into the following 12 groups: (1) chloroplast, (2) cytoplasm, (3) cytoskeleton, (4) endoplasmic reticulum, (5) extracellular, (6) Golgi apparatus, (7) lysosome, (8) mitochondria, (9) nucleus, (10) peroxisome, (11) plasma membrane, and (12) vacuole, which account for most organelles and subcellular compartments in an animal or plant cell. Examinations for self-consistency and jackknife testing of the SVMs method were conducted for three sets consisting of 1,911, 2,044, and 2,191 proteins. The correct rates for self-consistency and the jackknife test values achieved with these protein sets were 94 and 83% for 1,911 proteins, 92 and 78% for 2,044 proteins, and 89 and 75% for 2,191 proteins, respectively. Furthermore, tests for correct prediction rates were undertaken with three independent testing datasets containing 2,148 proteins, 2,417 proteins, and 2,494 proteins producing values of 84, 77, and 74%, respectively.     

Support vector machines for prediction of protein domain structural class

The support vector machines (SVMs) method was introduced for predicting the structural class of protein domains. The results obtained through the self-consistency test, jack-knife test, and independent dataset test have indicated that the current method and the elegant component-coupled algorithm developed by Chou and co-workers, if effectively complemented with each other, may become a powerful tool for predicting the structural class of protein domains.     

Esub8: A novel tool to predict protein subcellular localizations in eukaryotic organisms

Background

Subcellular localization of a new protein sequence is very important and fruitful for understanding its function. As the number of new genomes has dramatically increased over recent years, a reliable and efficient system to predict protein subcellular location is urgently needed.

Results

Esub8 was developed to predict protein subcellular localizations for eukaryotic proteins based on amino acid composition. In this research, the proteins are classified into the following eight groups: chloroplast, cytoplasm, extracellular, Golgi apparatus, lysosome, mitochondria, nucleus and peroxisome. We know subcellular localization is a typical classification problem; consequently, a one-against-one (1-v-1) multi-class support vector machine was introduced to construct the classifier. Unlike previous methods, ours considers the order information of protein sequences by a different method. Our method is tested in three subcellular localization predictions for prokaryotic proteins and four subcellular localization predictions for eukaryotic proteins on Reinhardt's dataset. The results are then compared to several other methods. The total prediction accuracies of two tests are both 100% by a self-consistency test, and are 92.9% and 84.14% by the jackknife test, respectively. Esub8 also provides excellent results: the total prediction accuracies are 100% by a self-consistency test and 87% by the jackknife test.

Conclusions

Our method represents a different approach for predicting protein subcellular localization and achieved a satisfactory result; furthermore, we believe Esub8 will be a useful tool for predicting protein subcellular localizations in eukaryotic organisms.

     

Using support vector machines to distinguish enzymes: Approached by incorporating wavelet transform

The enzymatic attributes of newly found protein sequences are usually determined either by biochemical analysis of eukaryotic and prokaryotic genomes or by microarray chips. These experimental methods are both time-consuming and costly. With the explosion of protein sequences registered in the databanks, it is highly desirable to develop an automated method to identify whether a given new sequence belongs to enzyme or non-enzyme. The discrete wavelet transform (DWT) and support vector machine (SVM) have been used in this study for distinguishing enzyme structures from non-enzymes. The networks have been trained and tested on two datasets of proteins with different wavelet basis functions, decomposition scales and hydrophobicity data types. Maximum accuracy has been obtained using SVM with a wavelet function of Bior2.4, a decomposition scale j=5, and Kyte-Doolittle hydrophobicity scales. The results obtained by the self-consistency test, jackknife test and independent dataset test are encouraging, which indicates that the proposed method can be employed as a useful assistant technique for distinguishing enzymes from non-enzymes.     

Prediction of protein structural class by amino acid and polypeptide composition.

A new approach of predicting structural classes of protein domain sequences is presented in this paper. Besides the amino acid composition, the composition of several dipeptides, tripeptides, tetrapeptides, pentapeptides and hexapeptides are taken into account based on the stepwise discriminant analysis. The result of jackknife test shows that this new approach can lead to higher predictive sensitivity and specificity for reduced sequence similarity datasets. Considering the dataset PDB40-B constructed by Brenner and colleagues, 75.2% protein domain sequences are correctly assigned in the jackknife test for the four structural classes: all-alpha, all-beta, alpha/beta and alpha + beta, which is improved by 19.4% in jackknife test and 25.5% in resubstitution test, in contrast with the component-coupled algorithm using amino acid composition alone (AAC approach) for the same dataset. In the cross-validation test with dataset PDB40-J constructed by Park and colleagues, more than 80% predictive accuracy is obtained. Furthermore, for the dataset constructed by Chou and Maggiona, the accuracy of 100% and 99.7% can be easily achieved, respectively, in the resubstitution test and in the jackknife test merely taking the composition of dipeptides into account. Therefore, this new method provides an effective tool to extract valuable information from protein sequences, which can be used for the systematic analysis of small or medium size protein sequences. The computer programs used in this paper are available on request.     

Support vector machines for the classification and prediction of beta-turn types.

The support vector machines (SVMs) method is proposed because it can reflect the sequence-coupling effect for a tetrapeptide in not only a beta-turn or non-beta-turn, but also in different types of beta-turn. The results of the model for 6022 tetrapeptides indicate that the rates of self-consistency for beta-turn types I, I', II, II', VI and VIII and non-beta-turns are 99.92%, 96.8%, 98.02%, 97.75%, 100%, 97.19% and 100%, respectively. Using these training data, the rate of correct prediction by the SVMs for a given protein: rubredoxin (54 residues. 51 tetrapeptides) which includes 12 beta-turn type I tetrapeptides, 1 beta-turn type II tetrapeptide and 38 non-beta-turns reached 82.4%. The high quality of prediction of the SVMs implies that the formation of different beta-turn types or non-beta-turns is considerably correlated with the sequence of a tetrapeptide. The SVMs can save CPU time and avoid the overfitting problem compared with the neural network method.     

Support vector machines for prediction of peptidyl prolyl cis/trans isomerization.

A new method for peptidyl prolyl cis/trans isomerization prediction based on the theory of support vector machines (SVM) was introduced. The SVM represents a new approach to supervised pattern classification and has been successfully applied to a wide range of pattern recognition problems. In this study, six training datasets consisting of different length local sequence respectively were used. The polynomial kernel functions with different parameter d were chosen. The test for the independent testing dataset and the jackknife test were both carried out. When the local sequence length was 20-residue and the parameter d = 8, the SVM method archived the best performance with the correct rate for the cis and trans forms reaching 70.4 and 69.7% for the independent testing dataset, 76.7 and 76.6% for the jackknife test, respectively. Matthew's correlation coefficients for the jackknife test could reach about 0.5. The results obtained through this study indicated that the SVM method would become a powerful tool for predicting peptidyl prolyl cis/trans isomerization.     

Prediction of rat protein subcellular localization with pseudo amino acid composition based on multiple sequential features

The study of rat proteins is an indispensable task in experimental medicine and drug development. The function of a rat protein is closely related to its subcellular location. Based on the above concept, we construct the benchmark rat proteins dataset and develop a combined approach for predicting the subcellular localization of rat proteins. From protein primary sequence, the multiple sequential features are obtained by using of discrete Fourier analysis, position conservation scoring function and increment of diversity, and these sequential features are selected as input parameters of the support vector machine. By the jackknife test, the overall success rate of prediction is 95.6% on the rat proteins dataset. Our method are performed on the apoptosis proteins dataset and the Gram-negative bacterial proteins dataset with the jackknife test, the overall success rates are 89.9% and 96.4%, respectively. The above results indicate that our proposed method is quite promising and may play a complementary role to the existing predictors in this area.     

Using ensemble classifier to identify membrane protein types

Predicting membrane protein type is both an important and challenging topic in current molecular and cellular biology. This is because knowledge of membrane protein type often provides useful clues for determining, or sheds light upon, the function of an uncharacterized membrane protein. With the explosion of newly-found protein sequences in the post-genomic era, it is in a great demand to develop a computational method for fast and reliably identifying the types of membrane proteins according to their primary sequences. In this paper, a novel classifier, the so-called "ensemble classifier", was introduced. It is formed by fusing a set of nearest neighbor (NN) classifiers, each of which is defined in a different pseudo amino acid composition space. The type for a query protein is determined by the outcome of voting among these constituent individual classifiers. It was demonstrated through the self-consistency test, jackknife test, and independent dataset test that the ensemble classifier outperformed other existing classifiers widely used in biological literatures. It is anticipated that the idea of ensemble classifier can also be used to improve the prediction quality in classifying other attributes of proteins according to their sequences.     

Support vector machines for predicting rRNA-, RNA-, and DNA-binding proteins from amino acid sequence

Classification of gene function remains one of the most important and demanding tasks in the post-genome era. Most of the current predictive computer methods rely on comparing features that are essentially linear to the protein sequence. However, features of a protein nonlinear to the sequence may also be predictive to its function. Machine learning methods, for instance the Support Vector Machines (SVMs), are particularly suitable for exploiting such features. In this work we introduce SVM and the pseudo-amino acid composition, a collection of nonlinear features extractable from protein sequence, to the field of protein function prediction. We have developed prototype SVMs for binary classification of rRNA-, RNA-, and DNA-binding proteins. Using a protein's amino acid composition and limited range correlation of hydrophobicity and solvent accessible surface area as input, each of the SVMs predicts whether the protein belongs to one of the three classes. In self-consistency and cross-validation tests, which measures the success of learning and prediction, respectively, the rRNA-binding SVM has consistently achieved >95% accuracy. The RNA- and DNA-binding SVMs demonstrate more diverse accuracy, ranging from approximately 76% to approximately 97%. Analysis of the test results suggests the directions of improving the SVMs.     

Predicting membrane protein types using residue-pair models based on reduced similarity dataset

An algorithm to predict the membrane protein types based on the multi-residue-pair effect in the Markov model is proposed. For a newly constructed dataset of 835 membrane proteins with very low sequence similarity, the overall prediction accuracy has been achieved as high as 81.1% and 71.7% in the resubstitution and jackknife test, respectively, for a prediction of type I single-pass, type II single-pass, multi-pass membrane proteins, lipid chain-anchored and GPI-anchored membrane proteins. The improvement of about 11% in the jackknife test can be achieved compared with the component-coupled algorithm merely based on the amino acid composition (AAC approach). The improvement is also confirmed on a high similarity dataset and the other extrapolating test. The result implies that designing more incisive analysis tools, one should develop algorithms based on the representative dataset with lower sequence similarity. The present algorithm is useful to expedite the determination of the types and functions of new membrane proteins and may be useful for the systematic analysis of functional genome data in a large scale. The computer program is available on request.     

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.     

Predicting enzyme family classes by hybridizing gene product composition and pseudo-amino acid composition

A new method has been developed to predict the enzymatic attribute of proteins by hybridizing the gene product composition and pseudo amino acid composition. As a demonstration, a working dataset was generated with a cutoff of 60% sequence identity to avoid redundancy and bias in statistical prediction. The dataset thus constructed contains 39989 protein sequences, of which 27469 are non-enzymes and 12520 enzymes that were further classified into 6 enzyme family classes according to their 6 main EC (Enzyme Commission) numbers (2314 are oxidoreductases, 3653 transferases, 3246 hydrolases, 1307 lyases, 676 isomerases, and 1324 ligases). The overall success rate by the jackknife test for the identification between enzyme and non-enzyme was 94%, and that for the identification among the 6 enzyme family classes was 98%. It is anticipated that, with the rapid increase of protein sequences entering into databanks, the current method will become a useful automated tool in identifying the enzymatic attribute of a newly found protein sequence.     

Predicting protease types by hybridizing gene ontology and pseudo amino acid composition

Proteases play a vitally important role in regulating most physiological processes. Different types of proteases perform different functions with different biological processes. Therefore, it is highly desired to develop a fast and reliable means to identify the types of proteases according to their sequences, or even just identify whether they are proteases or nonproteases. The avalanche of protein sequences generated in the postgenomic era has made such a challenge become even more critical and urgent. By hybridizing the gene ontology approach and pseudo amino acid composition approach, a powerful predictor called GO-PseAA predictor was introduced to address the problems. To avoid redundancy and bias, demonstrations were performed on a dataset where none of proteins has >/= 25% sequence identity to any other. The overall success rates thus obtained by the jackknife cross-validation test in identifying protease and nonprotease was 91.82%, and that in identifying the protease type was 85.49% among the following five types: (1) aspartic, (2) cysteine, (3) metallo, (4) serine, and (5) threonine. The high jackknife success rates yielded for such a stringent dataset indicate the GO-PseAA predictor is very powerful and might become a useful tool in bioinformatics and proteomics.     

Kernel-based machine learning protocol for predicting DNA-binding proteins

DNA-binding proteins (DNA-BPs) play a pivotal role in various intra- and extra-cellular activities ranging from DNA replication to gene expression control. Attempts have been made to identify DNA-BPs based on their sequence and structural information with moderate accuracy. Here we develop a machine learning protocol for the prediction of DNA-BPs where the classifier is Support Vector Machines (SVMs). Information used for classification is derived from characteristics that include surface and overall composition, overall charge and positive potential patches on the protein surface. In total 121 DNA-BPs and 238 non-binding proteins are used to build and evaluate the protocol. In self-consistency, accuracy value of 100% has been achieved. For cross-validation (CV) optimization over entire dataset, we report an accuracy of 90%. Using leave 1-pair holdout evaluation, the accuracy of 86.3% has been achieved. When we restrict the dataset to less than 20% sequence identity amongst the proteins, the holdout accuracy is achieved at 85.8%. Furthermore, seven DNA-BPs with unbounded structures are all correctly predicted. The current performances are better than results published previously. The higher accuracy value achieved here originates from two factors: the ability of the SVM to handle features that demonstrate a wide range of discriminatory power and, a different definition of the positive patch. Since our protocol does not lean on sequence or structural homology, it can be used to identify or predict proteins with DNA-binding function(s) regardless of their homology to the known ones.     

Using optimized evidence-theoretic K-nearest neighbor classifier and pseudo-amino acid composition to predict membrane protein types

Knowledge of membrane protein type often provides crucial hints toward determining the function of an uncharacterized membrane protein. With the avalanche of new protein sequences emerging during the post-genomic era, it is highly desirable to develop an automated method that can serve as a high throughput tool in identifying the types of newly found membrane proteins according to their primary sequences, so as to timely make the relevant annotations on them for the reference usage in both basic research and drug discovery. Based on the concept of pseudo-amino acid composition [K.C. Chou, Proteins: Struct. Funct. Genet. 43 (2001) 246-255; Erratum: Proteins: Struct. Funct. Genet. 44 (2001) 60] that has made it possible to incorporate a considerable amount of sequence-order effects by representing a protein sample in terms of a set of discrete numbers, a novel predictor, the so-called "optimized evidence-theoretic K-nearest neighbor" or "OET-KNN" classifier, was proposed. It was demonstrated via the self-consistency test, jackknife test, and independent dataset test that the new predictor, compared with many previous ones, yielded higher success rates in most cases. The new predictor can also be used to improve the prediction quality for, among many other protein attributes, structural class, subcellular localization, enzyme family class, and G-protein coupled receptor type. The OET-KNN classifier will be available as a web-server at http://www.pami.sjtu.edu.cn/kcchou.     

Prediction of membrane protein types and subcellular locations.

Membrane proteins are classified according to two different schemes. In scheme 1, they are discriminated among the following five types: (1) type I single-pass transmembrane, (2) type II single-pass transmembrane, (3) multipass transmembrane, (4) lipid chain-anchored membrane, and (5) GPI-anchored membrane proteins. In scheme 2, they are discriminated among the following nine locations: (1) chloroplast, (2) endoplasmic reticulum, (3) Golgi apparatus, (4) lysosome, (5) mitochondria, (6) nucleus, (7) peroxisome, (8) plasma, and (9) vacuole. An algorithm is formulated for predicting the type or location of a given membrane protein based on its amino acid composition. The overall rates of correct prediction thus obtained by both self-consistency and jackknife tests, as well as by an independent dataset test, were around 76-81% for the classification of five types, and 66-70% for the classification of nine cellular locations. Furthermore, classification and prediction were also conducted between inner and outer membrane proteins; the corresponding rates thus obtained were 88-91%. These results imply that the types of membrane proteins, as well as their cellular locations and other attributes, are closely correlated with their amino acid composition. It is anticipated that the classification schemes and prediction algorithm can expedite the functionality determination of new proteins. The concept and method can be also useful in the prioritization of genes and proteins identified by genomics efforts as potential molecular targets for drug design.