Novel hybrid method for the evaluation of parameters contributing in determination of protein structural classes

Due to the increasing gap between structure-determined and sequenced proteins, prediction of protein structural classes has been an important problem. It is very important to use efficient sequential parameters for developing class predictors because of the close sequence-structure relationship. The multinomial logistic regression model was used for the first time to evaluate the contribution of sequence parameters in determining the protein structural class. An in-house program generated parameters including single amino acid and all dipeptide composition frequencies. Then, the most effective parameters were selected by a multinomial logistic regression. Selected variables in the multinomial logistic model were Valine among single amino acid composition frequencies and Ala-Gly, Cys-Arg, Asp-Cys, Glu-Tyr, Gly-Glu, His-Tyr, Lys-Lys, Leu-Asp, Leu-Arg, Pro-Cys, Gln-Met, Gln-Thr, Ser-Trp, Val-Asn and Trp-Asn among dipeptide composition frequencies. Also a neural network model was constructed and fed by the parameters selected by multinomial logistic regression to build a hybrid predictor. In this study, self-consistency and jackknife tests on a database constructed by Zhou [1998. An intriguing controversy over protein structural class prediction. J. Protein Chem. 17(8), 729-738] containing 498 proteins are used to verify the performance of this hybrid method, and are compared with some of prior works. The results showed that our two-stage hybrid model approach is very promising and may play a complementary role to the existing powerful approaches.     

A neural network method for prediction of beta-turn types in proteins using evolutionary information

MOTIVATION: The prediction of beta-turns is an important element of protein secondary structure prediction. Recently, a highly accurate neural network based method Betatpred2 has been developed for predicting beta-turns in proteins using position-specific scoring matrices (PSSM) generated by PSI-BLAST and secondary structure information predicted by PSIPRED. However, the major limitation of Betatpred2 is that it predicts only beta-turn and non-beta-turn residues and does not provide any information of different beta-turn types. Thus, there is a need to predict beta-turn types using an approach based on multiple sequence alignment, which will be useful in overall tertiary structure prediction. RESULTS: In the present work, a method has been developed for the prediction of beta-turn types I, II, IV and VIII. For each turn type, two consecutive feed-forward back-propagation networks with a single hidden layer have been used where the first sequence-to-structure network has been trained on single sequences as well as on PSI-BLAST PSSM. The output from the first network along with PSIPRED predicted secondary structure has been used as input for the second-level structure-to-structure network. The networks have been trained and tested on a non-homologous dataset of 426 proteins chains by 7-fold cross-validation. It has been observed that the prediction performance for each turn type is improved significantly by using multiple sequence alignment. The performance has been further improved by using a second level structure-to-structure network and PSIPRED predicted secondary structure information. It has been observed that Type I and II beta-turns have better prediction performance than Type IV and VIII beta-turns. The final network yields an overall accuracy of 74.5, 93.5, 67.9 and 96.5% with MCC values of 0.29, 0.29, 0.23 and 0.02 for Type I, II, IV and VIII beta-turns, respectively, and is better than random prediction. AVAILABILITY: A web server for prediction of beta-turn types I, II, IV and VIII based on above approach is available at http://www.imtech.res.in/raghava/betaturns/ and http://bioinformatics.uams.edu/mirror/betaturns/ (mirror site).     

Optical spectroscopic elucidation of beta-turns in disulfide bridged cyclic tetrapeptides

Vibrational circular dichroism (VCD) spectroscopic features of type II beta-turns were characterized previously, but, criteria for differentiation between beta-turn types had not been established yet. Model tetrapeptides, cyclized through a disulfide bridge, were designed on the basis of previous experimental results and the observed incidence of amino acid residues in the i + 1 and i + 2 positions in beta-turns, to determine the features of VCD spectra of type I and II beta-turns. The results were correlated with electronic circular dichroism (ECD) spectra and VCD spectra calculated from conformational data obtained by molecular dynamics (MD) simulations. All cyclic tetrapeptides yielded VCD signals with a higher frequency negative and a lower frequency positive couplet with negative lobes overlapping. MD simulations confirmed the conformational homogeneity of these peptides in solution. Comparison with ECD spectroscopy, MD, and quantum chemical calculation results suggested that the low frequency component of VCD spectra originating from the tertiary amide vibrations could be used to distinguish between types of beta-turn structures. On the basis of this observation, VCD spectroscopic features of type II and VIII beta-turns and ECD spectroscopic properties of a type VIII beta-turn were suggested. The need for independent experimental as well as theoretical investigations to obtain decisive conformational information was recognized.     

Analysis and prediction of the different types of beta-turn in proteins

beta-Turns have been extracted from 59 non-identical proteins (resolution 2 A) using the standard criterion that the distance between C alpha (i) and C alpha (i + 3) is less than 7 A (1 A = 0.1 nm). The beta-turns have been classified, using phi, psi angles, into seven conventional turn types (I, I', II, II', IV, VIa, VIb) and a new class of beta-turn, designated type VIII, in which the central residues (i + 1, i + 2) adopt an alpha R beta conformation. Most beta-turn types are found in various topological environments, with the exception of I' and II' beta-turns, where 83% and 50%, respectively, are found in beta-hairpins. Sufficient data have been gathered to enable, for the first time, the separate statistical analysis of type I and II beta-turns. The two turn types have been shown to be strikingly different in their sequence preferences. Type I turns favour Asp, Asn, Ser and Cys at i; Asp, Ser, Thr and Pro at i + 1; Asp, Ser, Asn and Arg at i + 2; Gly, Trp and Met at i + 3, whilst type II turns prefer Pro at i + 1; Gly and Asn at i + 2; Gln and Arg at i + 3. These preferences have been explained by the specific side-chain interactions observed within the X-ray structures. The positional trends for type I and II beta-turns have been incorporated into the simple empirical predictive algorithm originally developed by P.N. Lewis et al. The program has improved the positional prediction of beta-turns, and has enhanced and extended the method by predicting the type of beta-turn. Since the observed preferences reflect local interactions these predictions are applicable not only to proteins, but also to peptides, many of which are thought to contain beta-turns.     

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.     

Vicinal disulfide bridge conformers by experimental methods and by ab initio and DFT molecular computations

A systematic comparison is made between experimental and computational data gained on vicinal disulfide bridges in proteins and peptides. Structural and stability data of ab initio and density functional theory (DFT) calculations on the model compound 4,5-ditiaheptano-7-lactam and the model peptide HCO-ox-[Cys-Cys]-NH2 at RHF/3-21G*, B3LYP/6-31+G(d), and B3LYP/6-311++G(d,p) levels of theory are presented. The data on Xxx-Cys-Cys-Yyy type amino acid sequence units retrieved from PDB SELECT, along with data on sequence units that have vicinal disulfide bridge, taken from the Brookhaven Protein Data Bank, are conformationally characterized. Amino acid backbone conformations, cis-trans isomerism of the amide bond between the two cysteine residues, and ring puckering are studied. Ring puckers are characterized by their relation to the conformers of the parent 4,5-ditiaheptano-7-lactam. Computational precision and accuracy are proved by frequency calculation and solvent model optimization on selected conformers. It is found that the ox-[Cys-Cys] unit is able to accept types I, II, VIa, VIb, and VIII beta-turn structures.     

Cyclic pentapeptides as models for reverse turns: determination of the equilibrium distribution between type I and type II conformations of Pro-Asn and Pro-Ala beta-turns

Cyclic pentapeptides are excellent models for reverse turns and have been used extensively in our laboratory to explore the influence of different amino acid sequences on turn preference. This paper is divided into two parts: In the first, we review our previous studies of cyclic pentapeptides. We summarize work that demonstrates the range of conformations possible within the cyclic pentapeptide backbone, the importance of sequence chirality in determining the backbone fold, and the utility of these cyclic pentapeptides as models for various turns. In the second, we present new results on two cyclic pentapeptides that contain beta-turns with Pro-Ala or Pro-Asn sequences in the i + 1 and i + 2 positions. By stereochemical criteria, a type I beta-turn is expected to be preferred by such L-L sequences. On the other hand, in proteins Asn occurs frequently in the i + 2 position of type II turns. We asked whether the same propensity would be manifest in an isolated model peptide, and if so, what the interactions were that influenced the relative stability of the type I and type II turns. To address these questions we have compared the conformational behavior of two peptides: cyclo(Gly-Pro-Ala-D-Phe-Pro) and cyclo(D-Ala-Pro-Asn-Gly-Pro). From previous studies, we anticipated that both peptides would contain an inverse gamma-turn and a beta-turn which consisted of either Gly-Pro-Ala-D-Phe or D-Ala-Pro-Asn-Gly in positions i to i + 3, respectively. Nuclear magnetic resonance analysis confirms this overall backbone conformation. Furthermore, quantitative nuclear Overhauser effect measurements in combination with molecular dynamics simulations and torsionally-forced energy minimizations have enabled us to determine that both type I and type II beta-turns are present in equilibrium in these peptides. The introduction of Asn in position i + 2 shifts this equilibrium significantly towards type II. We have done preliminary assessment of the possible side-chain/backbone conformations that contribute to the shift in populations.     

Analysis of Schizophrenia Data Using A Nonlinear Threshold Index Logistic Model

Genetic information, such as single nucleotide polymorphism (SNP) data, has been widely recognized as useful in prediction of disease risk. However, how to model the genetic data that is often categorical in disease class prediction is complex and challenging. In this paper, we propose a novel class of nonlinear threshold index logistic models to deal with the complex, nonlinear effects of categorical/discrete SNP covariates for Schizophrenia class prediction. A maximum likelihood methodology is suggested to estimate the unknown parameters in the models. Simulation studies demonstrate that the proposed methodology works viably well for moderate-size samples. The suggested approach is therefore applied to the analysis of the Schizophrenia classification by using a real set of SNP data from Western Australian Family Study of Schizophrenia (WAFSS). Our empirical findings provide evidence that the proposed nonlinear models well outperform the widely used linear and tree based logistic regression models in class prediction of schizophrenia risk with SNP data in terms of both Types I/II error rates and ROC curves.     

Configurationally driven folding of model tetrapeptides containing L- or D-morpholine-3-carboxylic acids as beta-turn nucleators

The conformational analysis by NMR, IR, and molecular modeling of tetrapeptides containing morpholine-3-carboxylic acid (Mor) as a proline surrogate is presented. The relationship between the chirality of the cyclic amino acid at position i+1 and the turn propensity is maintained with respect to the reference proline-containing peptides, although marked differences in the type of folded structures were observed. The conformational profile of morpholine-containing turn peptides as a function of the chirality of the cyclic amino acid indicated that the heterochiral tetrapeptide containing the D-isomer of the cyclic amino acid is more prone to nucleate compact folded structures, although with no resemblance to the beta-turn structures of D-proline-containing peptides. Also, the solvation system proved to influence the organization of folded structures, as in the more interactive CD(3)CN the model peptides showed more compact conformations. The L-Mor-containing peptide displayed two rotamers at the Val-Mor amide bond. The trans isomer did not experience any turn structures, nor any intramolecular hydrogen-bonds, whereas the cis isomer showed a strong preference for a type VI beta-turn structure, thus providing a different conformational asset with respect to the beta-turn structure as reported for the reference L-proline model peptide.     

Role of azaamino acid residue in beta-turn formation and stability in designed peptide.

The structural perturbation induced by C(alpha)-->N(alpha) exchange in azaamino acid-containing peptides was predicted by ab initio calculation of the 6-31G* and 3-21G* levels. The global energy-minimum conformations for model compounds, For-azaXaa-NH2 (Xaa=Gly, Ala, Leu) appeared to be the beta-turn motif with a dihedral angle of phi= +/- 90 degrees, psi=0 degrees. This suggests that incorporation of the azaXaa residue into the i+2 position of designed peptides could stabilize the beta-turn structure. The model azaLeu-containing peptide, Boc-Phe-azaLeu-Ala-OMe, which is predicted to adopt a beta-turn conformation was designed and synthesized in order to experimentally elucidate the role of the azaamino acid residue. Its structural preference in organic solvents was investigated using 1H NMR, molecular modelling and IR spectroscopy. The temperature coefficients of amide protons, the characteristic NOE patterns, the restrained molecular dynamics simulation and IR spectroscopy defined the dihedral angles [ (phi i+1, psi i+1) (phi i+2, psi i+2)] of the Phe-azaLeu fragment in the model peptide, Boc-Phe-azaLeu-Ala-OMe, as [(-59 degrees, 127 degrees) (107 degrees, -4 degrees)]. This solution conformation supports a betaII-turn structural preference in azaLeu-containing peptides as predicted by the quantum chemical calculation. Therefore, intercalation of the azaamino acid residue into the i+2 position in synthetic peptides is expected to provide a stable beta-turn formation, and this could be utilized in the design of new peptidomimetics adopting a beta-turn scaffold.     

NetTurnP--neural network prediction of beta-turns by use of evolutionary information and predicted protein sequence features

β-turns are the most common type of non-repetitive structures, and constitute on average 25% of the amino acids in proteins. The formation of β-turns plays an important role in protein folding, protein stability and molecular recognition processes. In this work we present the neural network method NetTurnP, for prediction of two-class β-turns and prediction of the individual β-turn types, by use of evolutionary information and predicted protein sequence features. It has been evaluated against a commonly used dataset BT426, and achieves a Matthews correlation coefficient of 0.50, which is the highest reported performance on a two-class prediction of β-turn and not-β-turn. Furthermore NetTurnP shows improved performance on some of the specific β-turn types. In the present work, neural network methods have been trained to predict β-turn or not and individual β-turn types from the primary amino acid sequence. The individual β-turn types I, I', II, II', VIII, VIa1, VIa2, VIba and IV have been predicted based on classifications by PROMOTIF, and the two-class prediction of β-turn or not is a superset comprised of all β-turn types. The performance is evaluated using a golden set of non-homologous sequences known as BT426. Our two-class prediction method achieves a performance of: MCC=0.50, Qtotal=82.1%, sensitivity=75.6%, PPV=68.8% and AUC=0.864. We have compared our performance to eleven other prediction methods that obtain Matthews correlation coefficients in the range of 0.17-0.47. For the type specific β-turn predictions, only type I and II can be predicted with reasonable Matthews correlation coefficients, where we obtain performance values of 0.36 and 0.31, respectively. CONCLUSION: The NetTurnP method has been implemented as a webserver, which is freely available at http://www.cbs.dtu.dk/services/NetTurnP/. NetTurnP is the only available webserver that allows submission of multiple sequences.     

A comparative cytogenetic study of the rice (<Emphasis Type="Italic">Oryza sativa</Emphasis> L.) autotetraploid restorers and hybrids

We studied pollen fertility, seed set and cytogenetic characteristics of restorer lines and F1 hybrids of autotetraploid rice. T4002, T4063, T461A × T4002 and T461A × T4063 showed significantly higher pollen fertility and seed set than T4132 and T461A × T4132. Meiotic pairing configurations of T4002, T4063, T4132, T461A × T4002, T461A × T4063 and T461A × T4132 were 0.05 I + 19.96 II (9.89 rod + 10.07 ring) + 0.01 III + 2.00 IV, 0.11 I + 19.17 II (8.90 rod + 10.37 ring) + 0.09 III + 2.26 IV + 0.01 VI, 1.34 I + 9.46 II (4.50 rod + 4.96 ring) + 0.80 III + 6.02 IV + 0.09 VI + 0.09 VIII, 0.02 I + 14.36 II (6.44 rod + 7.91 ring) + 0.01 III + 4.80 IV + 0.01 VIII, 0.06 I + 17.67 II (11.01 rod + 6.67 ring) + 0.06 III + 3.10 IV + 0.01 VI and 1.11 I + 11.31 II (5.80 rod + 5.51 ring) + 0.41 III + 5.63 IV + 0.03 VI + 0.03 VIII, respectively. Configuration 16 II + 4 IV and 12 II + 6 IV occurred in the highest frequency among the autotetraploid restorers and hybrids. Meiotic chromosome behaviors were less abnormal in the tetraploids with high seed set than those with low seed set. The hybrids had fewer frequencies of bivalents, univalents, trivalents and multivalents than the restorers, but higher frequency of quatrivalents than the restorers at MI. The frequency of univalents at MI had the most impact on pollen fertility and seed set, i.e., pollen fertility decreased with the increase of univalents. The secondary impact factors were trivalents and multivalents, and bivalents and quatrivalents had no effect on pollen fertility and seed set. The correlative relationship between pollen fertility and cytogenetic behaviors could be utilized to improve seed set in autotetraploidy breeding.     

Prediction of beta-turns from amino acid sequences using the residue-coupled model.

We evaluated the prediction of beta-turns from amino acid sequences using the residue-coupled model with an enlarged representative protein data set selected from the Protein Data Bank. Our results show that the probability values derived from a data set comprising 425 protein chains yielded an overall beta-turn prediction accuracy 68.74%, compared with 94.7% reported earlier on a data set of 30 proteins using the same method. However, we noted that the overall beta-turn prediction accuracy using probability values derived from the 30-protein data set reduces to 40.74% when tested on the data set comprising 425 protein chains. In contrast, using probability values derived from the 425 data set used in this analysis, the overall beta-turn prediction accuracy yielded consistent results when tested on either the 30-protein data set (64.62%) used earlier or a more recent representative data set comprising 619 protein chains (64.66%) or on a jackknife data set comprising 476 representative protein chains (63.38%). We therefore recommend the use of probability values derived from the 425 representative protein chains data set reported here, which gives more realistic and consistent predictions of beta-turns from amino acid sequences.     

An analysis of the secondary structure of spider spidroins I and II belonging to different species

We have analyzed the secondary structure of spidroin proteins of I and II types, related to spiders of different species. We used standard methods of secondary structure prediction NNPREDICT and JPRED and also analyzed the occurrences of oligopeptides with a preferred secondary structure with the help of the OLIGON program. We have demonstrated that local segments of the polypeptide chain can adopt alpha- and beta-conformations as well as the left-handed helix of polyproline II type. Periodical patterns found in the amino acid distribution indicate that there is a possibility of development of a macroscopic order accompanied by local conformational transitions.     

BTEVAL: a server for evaluation of beta-turn prediction methods

This paper describes a web server BTEVAL, developed for assessing the performance of newly developed beta-turn prediction method and it's ranking with respect to other existing beta-turn prediction methods. Evaluation of a method can be carried out on a single protein or a number of proteins. It consists of clean data set of 426 non-homologous proteins with seven subsets of these proteins. Users can evaluate their method on any subset or a complete set of data. The method is assessed at amino acid level and performance is evaluated in terms of Qtotal, Qpredicted, Qobserved and MCC measures. The server also compares the performance of the method with other existing beta-turn prediction methods such as Chou-Fasman algorithm, Thornton's algorithm, GORBTURN, 1-4 and 2-3 Correlation model, Sequence coupled model and BTPRED. The server is accessible from http://imtech.res.in/raghava/bteval/     

Support vector machine-based method for subcellular localization of human proteins using amino acid compositions, their order, and similarity search

Here we report a systematic approach for predicting subcellular localization (cytoplasm, mitochondrial, nuclear, and plasma membrane) of human proteins. First, support vector machine (SVM)-based modules for predicting subcellular localization using traditional amino acid and dipeptide (i + 1) composition achieved overall accuracy of 76.6 and 77.8%, respectively. PSI-BLAST, when carried out using a similarity-based search against a nonredundant data base of experimentally annotated proteins, yielded 73.3% accuracy. To gain further insight, a hybrid module (hybrid1) was developed based on amino acid composition, dipeptide composition, and similarity information and attained better accuracy of 84.9%. In addition, SVM modules based on a different higher order dipeptide i.e. i + 2, i + 3, and i + 4 were also constructed for the prediction of subcellular localization of human proteins, and overall accuracy of 79.7, 77.5, and 77.1% was accomplished, respectively. Furthermore, another SVM module hybrid2 was developed using traditional dipeptide (i + 1) and higher order dipeptide (i + 2, i + 3, and i + 4) compositions, which gave an overall accuracy of 81.3%. We also developed SVM module hybrid3 based on amino acid composition, traditional and higher order dipeptide compositions, and PSI-BLAST output and achieved an overall accuracy of 84.4%. A Web server HSLPred (www.imtech.res.in/raghava/hslpred/ or bioinformatics.uams.edu/raghava/hslpred/) has been designed to predict subcellular localization of human proteins using the above approaches.     

Prediction of the location and type of beta-turns in proteins using neural networks.

A neural network has been used to predict both the location and the type of beta-turns in a set of 300 nonhomologous protein domains. A substantial improvement in prediction accuracy compared with previous methods has been achieved by incorporating secondary structure information in the input data. The total percentage of residues correctly classified as beta-turn or not-beta-turn is around 75% with predicted secondary structure information. More significantly, the method gives a Matthews correlation coefficient (MCC) of around 0.35, compared with a typical MCC of around 0.20 using other beta-turn prediction methods. Our method also distinguishes the two most numerous and well-defined types of beta-turn, types I and II, with a significant level of accuracy (MCCs 0.22 and 0.26, respectively).     

Peptoid residues and beta-turn formation.

A set of terminally protected tripeptoids containing a residue of either N-methylglycine or N-isobutylglycine in position i + 1/i + 2 were synthesized and tested for intramolecularly H-bonded beta-turn formation. By exploiting FT-IR absorption and 1H NMR techniques, their folding tendencies were compared with those of a variety of reference peptides. The amount of beta-turn induction and the relative extent of the various types of intramolecularly H-bonded beta-turn conformers were determined in chloroform solution.