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.     

Mimicry by asx- and ST-turns of the four main types of beta-turn in proteins

Hydrogen-bonded beta-turns in proteins occur in four categories: type I (the most common), type II, type II', and type I'. Asx-turns resemble beta-turns, in that both have an NH. . .OC hydrogen bond forming a ring of 10 atoms. Serine and threonine side chains also commonly form hydrogen-bonded turns, here called ST-turns. Asx-turns and ST-turns can be categorized into four classes, based on side chain rotamers and the conformation of the central turn residue, which are geometrically equivalent to the four types of beta-turns. We propose asx- and ST-turns be named using the type I, II, I', and II' beta-turn nomenclature. Using this, the frequency of occurrence of both asx- and ST-turns is: type II' > type I > type II > type I', whereas for beta-turns it is type I > type II > type I' > type II'. Almost all type II asx-turns occur as a recently described three residue feature named an asx-nest.     

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).     

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.     

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).     

Effects of turn residues in directing the formation of the beta-sheet and in the stability of the beta-sheet

The designed peptide (denoted 20-mer, sequence VFITS(D)PGKTYTEV(D)PGOKILQ) has been shown to form a three-strand antiparallel beta-sheet. It is generally believed that the (D)Pro-Gly segment has the propensity to adopt a type II' beta-turn, thereby promoting the formation of this beta-sheet. Here, we replaced (D)Pro-Gly with Asp-Gly, which should favor a type I' turn, to examine the influence of different type of turns on the stability of the beta-sheet. Contrary to our expectation, the mutant peptide, denoted P6D, forms a five-residue type I turn plus a beta-bulge between the first two strands due to a one amino-acid frameshift in the hydrogen bonding network and side-chain inversion of the first beta-strand. In contrast, the same kind of substitution at (D)Pro-14 in the double mutant, denoted P6DP14D, does not yield the same effect. These observations suggest that the SDGK sequence disfavors the type I' conformation while the VDGO sequence favors a type I' turn, and that the frameshift in the first strand provides a way for the peptide to accommodate a disfavored turn sequence by protruding a bulge in the formation of the beta-hairpin. Thus, different types of turns can affect the stability of a beta-structure.     

Structural engineering of the HIV-1 protease molecule with a beta-turn mimic of fixed geometry.

An important goal in the de novo design of enzymes is the control of molecular geometry. To this end, an analog of the protease from human immunodeficiency virus 1 (HIV-1 protease) was prepared by total chemical synthesis, containing a constrained, nonpeptidic type II' beta-turn mimic of predetermined three-dimensional structure. The mimic beta-turn replaced residues Gly16,17 in each subunit of the homodimeric molecule. These residues constitute the central amino acids of two symmetry-related type I' beta-turns in the native, unliganded enzyme. The beta-turn mimic-containing enzyme analog was fully active, possessed the same substrate specificity as the Gly16,17-containing enzyme, and showed enhanced resistance to thermal inactivation. These results indicate that the precise geometry of the beta-turn at residues 15-18 in each subunit is not critical for activity, and that replacement of the native sequence with a rigid beta-turn mimic can lead to enhanced protein stability. Finally, the successful incorporation of a fixed element of secondary structure illustrates the potential of a "molecular kit set" approach to protein design and synthesis.     

A bioactive somatostatin analog without a type II' beta-turn: synthesis and conformational analysis in solution.

A cyclic somatostatin analog [structure: see text] (1) has been synthesized. Biological assays show that this compound has strong binding affinities to somatostatin hsst2 and hsst5 receptor subtypes (5.2 and 1.2 nM, respectively, and modest affinity to hsst4 (41.1 nM)). Our conformational analysis carried out in DMSO-d6 indicates that this compound exists as two structures arising from the trans and cis configurations of the peptide bond between Phe7 and N-alkylated Gly8. However, neither conformer exhibits a type II' beta-turn. This is the first report of a potent bioactive somatostatin analog that does not exhibit a type II' beta-turn in solution. Molecular dynamics simulations (500 ps) carried out at 300 K indicate that the backbone of compound 1 is more flexible than other cyclic somatostatin analogs formed by disulfide bonds.     

Prediction of beta-turns in proteins from multiple alignment using neural network

A neural network-based method has been developed for the prediction of beta-turns in proteins by using multiple sequence alignment. Two feed-forward back-propagation networks with a single hidden layer are used where the first-sequence structure network is trained with the multiple sequence alignment in the form of PSI-BLAST-generated position-specific scoring matrices. The initial predictions from the first network and PSIPRED-predicted secondary structure are used as input to the second structure-structure network to refine the predictions obtained from the first net. A significant improvement in prediction accuracy has been achieved by using evolutionary information contained in the multiple sequence alignment. The final network yields an overall prediction accuracy of 75.5% when tested by sevenfold cross-validation on a set of 426 nonhomologous protein chains. The corresponding Q(pred), Q(obs), and Matthews correlation coefficient values are 49.8%, 72.3%, and 0.43, respectively, and are the best among all the previously published beta-turn prediction methods. The Web server BetaTPred2 (http://www.imtech.res.in/raghava/betatpred2/) has been developed based on this approach.     

Designed beta-hairpin peptides with defined tight turn stereochemistry

The conformational analysis of two synthetic octapeptides, Boc-Leu-Val-Val-D-Pro-L-Ala-Leu-Val-Val-OMe (1) and Boc-Leu-Val-Val-D-Pro-D-Ala-Leu-Val-Val-OMe (2) has been carried out in order to investigate the effect of beta-turn stereochemistry on designed beta-hairpin structures. Five hundred megahertz (1)H NMR studies establish that both peptides 1 and 2 adopt predominantly beta-hairpin conformations in methanol solution. Specific nuclear Overhauser effects provide evidence for a type II' beta-turn conformation for the D-Pro-L-Ala segment in 1, while the NMR data suggest that the type I' D-Pro-D-Ala beta-turn conformation predominates in peptide 2. Evidence for a minor conformation in peptide 2, in slow exchange on the NMR time scale, is also presented. Interstrand registry is demonstrated in both peptides 1 and 2. The crystal structure of 1 reveals two independent molecules in the crystallographic asymmetric unit, both of which adopt beta-hairpin conformations nucleated by D-Pro-L-Ala type II' beta-turns and are stabilized by three cross-strand hydrogen bonds. CD spectra for peptides 1 and 2 show marked differences, presumably as a consequence of the superposition of spectral bands arising from both beta-turn and beta-strand conformations.     

Analysis and identification of beta-turn types using multinomial logistic regression and artificial neural network

MOTIVATION: So far various statistical and machine learning techniques applied for prediction of beta-turns. The majority of these techniques have been only focused on the prediction of beta-turn location in proteins. We developed a hybrid approach for analysis and prediction of different types of beta-turn. RESULTS: A two-stage hybrid model developed to predict the beta-turn Types I, II, IV and VIII. Multinomial logistic regression was initially used for the first time to select significant parameters in prediction of beta-turn types using a self-consistency test procedure. The extracted parameters were consisted of 80 amino acid positional occurrences and 20 amino acid percentages in beta-turn sequence. The most significant parameters were then selected using multinomial logistic regression model. Among these, the occurrences of glutamine, histidine, glutamic acid and arginine, respectively, in positions i, i + 1, i + 2 and i + 3 of beta-turn sequence had an overall relationship with five beta-turn types. A neural network model was then constructed and fed by the parameters selected by multinomial logistic regression to build a hybrid predictor. The networks have been trained and tested on a non-homologous dataset of 565 protein chains by 9-fold cross-validation. It has been observed that the hybrid model gives a Matthews correlation coefficient (MCC) of 0.235, 0.473, 0.103 and 0.124, respectively, for beta-turn Types I, II, IV and VIII. Our model also distinguished the different types of beta-turn in the embedded binary logit comparisons which have not carried out so far. AVAILABILITY: Available on request from the authors.     

Improved method for predicting beta-turn using support vector machine

MOTIVATION: Numerous methods for predicting beta-turns in proteins have been developed based on various computational schemes. Here, we introduce a new method of beta-turn prediction that uses the support vector machine (SVM) algorithm together with predicted secondary structure information. Various parameters from the SVM have been adjusted to achieve optimal prediction performance. RESULTS: The SVM method achieved excellent performance as measured by the Matthews correlation coefficient (MCC = 0.45) using a 7-fold cross validation on a database of 426 non-homologous protein chains. To our best knowledge, this MCC value is the highest achieved so far for predicting beta-turn. The overall prediction accuracy Qtotal was 77.3%, which is the best among the existing prediction methods. Among its unique attractive features, the present SVM method avoids overtraining and compresses information and provides a predicted reliability index.     

Preferred conformations of RGDX tetrapeptides to inhibit the binding of fibrinogen to platelets

The conformational study on Arg-Gly-Asp (RGD)-containing tetrapeptides in the unhydrated and hydrated states has been carried out using the force field ECEPP/3 and the hydration shell model. The tetrapeptides studied here are H-RGDX-OH (X = Trp, Tyr, Phe, Leu, Val, Cys, Gln, and Ser), which show the inhibitory activity for binding of fibrinogen to platelets in the order of RGDW approximately equal to RGDY approximately equal to RGDF approximately equal to RGDL > RGDV > or = RGDC > or = RGDQ > or = RGDS. The backbone conformations with two C(7) backbone-to-backbone hydrogen bonds between Asp and Arg residues and between Xaa and Gly residues are in common most probable for the RGD sequence of RGDX tetrapeptides in the hydrated state. The dominant beta-turns for RGDX are found to be the types V' and IV at Gly-Asp and Asp-Xaa sequences, respectively, which are quite similar to the types II' and I (or II), respectively. However, it cannot be ruled out that the extended conformations are also remarkably feasible for RGDX tetrapeptides in water by peering the distributions of backbone conformations. These calculated results are consistent with the experimental results on RGD-containing proteins and conformationally constrained RGD-containing peptides. The reason why the RGDX becomes more potent as the side chain of the X residue is more hydrophobic may be ascribed to that the more hydrophobic is the residue X, the more populated are beta-turn structures for the Gly-Asp sequence. The hydrophobic side chain of X residue exposed to water is likely to interact with the hydrophobic region of receptor easily.     

BetaTPred: prediction of beta-TURNS in a protein using statistical algorithms

MOTIVATION: beta-turns play an important role from a structural and functional point of view. beta-turns are the most common type of non-repetitive structures in proteins and comprise on average, 25% of the residues. In the past numerous methods have been developed to predict beta-turns in a protein. Most of these prediction methods are based on statistical approaches. In order to utilize the full potential of these methods, there is a need to develop a web server. RESULTS: This paper describes a web server called BetaTPred, developed for predicting beta-TURNS in a protein from its amino acid sequence. BetaTPred allows the user to predict turns in a protein using existing statistical algorithms. It also allows to predict different types of beta-TURNS e.g. type I, I', II, II', VI, VIII and non-specific. This server assists the users in predicting the consensus beta-TURNS in a protein. AVAILABILITY: The server is accessible from http://imtech.res.in/raghava/betatpred/     

PEPstr: a de novo method for tertiary structure prediction of small bioactive peptides

Among secondary structure elements, beta-turns are ubiquitous and major feature of bioactive peptides. We analyzed 77 biologically active peptides with length varying from 9 to 20 residues. Out of 77 peptides, 58 peptides were found to contain at least one beta-turn. Further, at the residue level, 34.9% of total peptide residues were found to be in beta-turns, higher than the number of helical (32.3%) and beta-sheet residues (6.9%). So, we utilized the predicted beta-turns information to develop an improved method for predicting the three-dimensional (3D) structure of small peptides. In principle, we built four different structural models for each peptide. The first 'model I' was built by assigning all the peptide residues an extended conformation (phi = Psi = 180 degrees ). Second 'model II' was built using the information of regular secondary structures (helices, beta-strands and coil) predicted from PSIPRED. In third 'model III', secondary structure information including beta-turn types predicted from BetaTurns method was used. The fourth 'model IV' had main-chain phi, Psi angles of model III and side chain angles assigned using standard Dunbrack backbone dependent rotamer library. These models were further refined using AMBER package and the resultant C(alpha) rmsd values were calculated. It was found that adding the beta-turns to the regular secondary structures greatly reduces the rmsd values both before and after the energy minimization. Hence, the results indicate that regular and irregular secondary structures, particularly beta-turns information can provide valuable and vital information in the tertiary structure prediction of small bioactive peptides. Based on the above study, a web server PEPstr (http://www.imtech.res.in/raghava/pepstr/) was developed for predicting the tertiary structure of small bioactive peptides.     

Quantitative analysis of cyclic beta-turn models.

The beta-turn is a frequently found structural unit in the conformation of globular proteins. Although the circular dichroism (CD) spectra of the alpha-helix and beta-pleated sheet are well defined, there remains some ambiguity concerning the pure component CD spectra of the different types of beta-turns. Recently, it has been reported (Hollósi, M., Kövér, K.E., Holly, S., Radics, L., & Fasman, G.D., 1987, Biopolymers 26, 1527-1572; Perczel, A., Hollósi, M., Foxman, B.M., & Fasman, G.D., 1991a, J. Am. Chem. Soc. 113, 9772-9784) that some pseudohexapeptides (e.g., the cyclo[(delta)Ava-Gly-Pro-Aaa-Gly] where Aaa = Ser, Ser(OtBu), or Gly) in many solvents adopt a conformational mixture of type I and the type II beta-turns, although the X-ray-determined conformation was an ideal type I beta-turn. In addition to these pseudohexapeptides, conformational analysis was also carried out on three pseudotetrapeptides and three pseudooctapeptides. The target of the conformation analysis reported herein was to determine whether the ring stress of the above beta-turn models has an influence on their conformational properties. Quantitative nuclear Overhauser effect (NOE) measurements yielded interproton distances. The conformational average distances so obtained were interpreted utilizing molecular dynamics (MD) simulations to yield the conformational percentages. These conformational ratios were correlated with the conformational weights obtained by quantitative CD analysis of the same compounds. The pure component CD curves of type I and type II beta-turns were also obtained, using a recently developed algorithm (Perczel, A., Tusnády, G., Hollósi, M., & Fasman, G.D., 1991b, Protein Eng. 4(6), 669-679). For the first time the results of a CD deconvolution, based on the CD spectra of 14 beta-turn models, were assigned by quantitative NOE results. The NOE experiments confirmed the ratios of the component curves found for the two major beta-turns by CD analysis. These results can now be used to enhance the conformational determination of globular proteins on the basis of their CD spectra.     

Structural models of the evolutionarily conservative central domain of silk-moth chorion proteins

Silk-moth chorion proteins belong to a small number of families: A, B, C, Hc-A and Hc-B. The central domain is an evolutionarily conservative region in each family, of variable length and composition between families. This domain shows clear 6-fold periodicities for various amino acid residues, e.g. glycine. The periodicities, together with the well-documented prevalence of beta-sheet and beta-turn secondary structure of chorion proteins, strongly support a structural model in which four-residue beta-strands alternate with beta-turns, forming a compact antiparallel, probably twisted beta-sheet. Conformational analysis, aided by interactive graphics refinement and recent experimental findings, further suggest that this structure consists of beta-strands, alternating with I' and II' beta-turns, and apparently forms the basis for the molecular and supramolecular assembly of chorion.     

Better prediction of the location of alpha-turns in proteins with support vector machine

We have developed a novel method named AlphaTurn to predict alpha-turns in proteins based on the support vector machine (SVM). The prediction was done on a data set of 469 nonhomologous proteins containing 967 alpha-turns. A great improvement in prediction performance was achieved by using multiple sequence alignment generated by PSI-BLAST as input instead of the single amino acid sequence. The introduction of secondary structure information predicted by PSIPRED also improved the prediction performance. Moreover, we handled the very uneven data set by combining the cost factor j with the "state-shifting" rule. This further promoted the prediction quality of our method. The final SVM model yielded a Matthews correlation coefficient (MCC) of 0.25 by a 10-fold cross-validation. To our knowledge, this MCC value is the highest obtained so far for predicting alpha-turns. An online Web server based on this method has been developed and can be freely accessed at http://bmc.hust.edu.cn/bioinformatics/ or http://210.42.106.80/.     

Influence of N-terminal residue stereochemistry on the prolyl amide geometry and the conformation of 5-tert-butylproline type VI beta-turn mimics.

The effects of N-terminal amino acid stereochemistry on prolyl amide geometry and peptide turn conformation were investigated by coupling both L- and D-amino acids to (2S, 5R)-5-tert-butylproline and L-proline to generate, respectively, N-(acetyl)dipeptide N'-methylamides 1 and 2. Prolyl amide cis- and trans-isomers were, respectively, favored for peptides 1 and 2 as observed by proton NMR spectroscopy in water, DMSO and chloroform. The influence of solvent composition on amide proton chemical shift indicated an intramolecular hydrogen bond between the N'-methylamide proton and the acetamide carbonyl for the major conformer of dipeptides (S)-1, that became less favorable in (R)-1 and 2. The coupling constant (3J(NH,alpha)) values for the cis-isomer of (R)-1 indicated a phi2 dihedral angle value characteristic of a type VIb beta-turn conformation in solution. X-ray crystallographic analysis of N-acetyl-D-leucyl-5-tert-butylproline N'-methylamide (R)-lb showed the prolyl residue in a type VIb beta-turn geometry possessing an amide cis-isomer and psi3-dihedral angle having a value of 157 degrees, which precluded an intramolecular hydrogen bond. Intermolecular hydrogen bonding between the leucyl residues of two turn structures within the unit cell positioned the N-terminal residue in a geometry where their phi2 and psi2 dihedral angle values were not characteristic of an ideal type VIb turn. The circular dichroism spectra of tert-butylprolyl peptides (S)- and (R)-1b were found not to be influenced by changes in solvent composition from water to acetonitrile. The type B spectrum exhibited by (S)-1b has been previously assigned to a type VIa beta-turn conformation [Halab L, Lubell WD. J. Org. Chem. 1999; 64: 3312-3321]. The type C spectrum exhibited by the (R)-lb has previously been associated with type II' beta-turn and alpha-helical conformations in solution and appears now to be also characteristic for a type VIb geometry.