An accurate feature-based method for identifying DNA-binding residues on protein surfaces

Proteins that interact with DNA play vital roles in all mechanisms of gene expression and regulation. In order to understand these activities, it is crucial to analyze and identify DNA-binding residues on DNA-binding protein surfaces. Here, we proposed two novel features B-factor and packing density in combination with several conventional features to characterize the DNA-binding residues in a well-constructed representative dataset of 119 protein-DNA complexes from the Protein Data Bank (PDB). Based on the selected features, a prediction model for DNA-binding residues was constructed using support vector machine (SVM). The predictor was evaluated using a 5-fold cross validation on above dataset of 123 DNA-binding proteins. Moreover, two independent datasets of 83 DNA-bound protein structures and their corresponding DNA-free forms were compiled. The B-factor and packing density features were statistically analyzed on these 83 pairs of holo-apo proteins structures. Finally, we developed the SVM model to accurately predict DNA-binding residues on protein surface, given the DNA-free structure of a protein. Results showed here indicate that our method represents a significant improvement of previously existing approaches such as DISPLAR. The observation suggests that our method will be useful in studying protein-DNA interactions to guide consequent works such as site-directed mutagenesis and protein-DNA docking.     

On the relation between residue flexibility and residue interactions in proteins

B-factor from X-ray crystal structure can well measure protein structural flexibility, which plays an important role in different biological processes, such as catalysis, binding and molecular recognition. Understanding the essence of flexibility can be helpful for the further study of the protein function. In this study, we attempted to correlate the flexibility of a residue to its interactions with other residues by representing the protein structure as a residue contact network. Here, several well established network topological parameters were employed to feature such interactions. A prediction model was constructed for B-factor of a residue by using support vector regression (SVR). Pearson correlation coefficient (CC) was used as the performance measure. CC values were 0.63 and 0.62 for single amino acid and for the whole sequence, respectively. Our results revealed well correlations between B-factors and network topological parameters. This suggests that the protein structural flexibility could be well characterized by the inter-amino acid interactions in a protein.     

Potential for assessing quality of protein structure based on contact number prediction

We developed a novel knowledge-based residue environment potential for assessing the quality of protein structures in protein structure prediction. The potential uses the contact number of residues in a protein structure and the absolute contact number of residues predicted from its amino acid sequence using a new prediction method based on a support vector regression (SVR). The contact number of an amino acid residue in a protein structure is defined by the number of residues around a given residue. First, the contact number of each residue is predicted using SVR from an amino acid sequence of a target protein. Then, the potential of the protein structure is calculated from the probability distribution of the native contact numbers corresponding to the predicted ones. The performance of this potential is compared with other score functions using decoy structures to identify both native structure from other structures and near-native structures from nonnative structures. This potential improves not only the ability to identify native structures from other structures but also the ability to discriminate near-native structures from nonnative structures.     

Protein flexibility and rigidity predicted from sequence

Structural flexibility has been associated with various biological processes such as molecular recognition and catalytic activity. In silico studies of protein flexibility have attempted to characterize and predict flexible regions based on simple principles. B-values derived from experimental data are widely used to measure residue flexibility. Here, we present the most comprehensive large-scale analysis of B-values. We used this analysis to develop a neural network-based method that predicts flexible-rigid residues from amino acid sequence. The system uses both global and local information (i.e., features from the entire protein such as secondary structure composition, protein length, and fraction of surface residues, and features from a local window of sequence-consecutive residues). The most important local feature was the evolutionary exchange profile reflecting sequence conservation in a family of related proteins. To illustrate its potential, we applied our method to 4 different case studies, each of which related our predictions to aspects of function. The first 2 were the prediction of regions that undergo conformational switches upon environmental changes (switch II region in Ras) and the prediction of surface regions, the rigidity of which is crucial for their function (tunnel in propeller folds). Both were correctly captured by our method. The third study established that residues in active sites of enzymes are predicted by our method to have unexpectedly low B-values. The final study demonstrated how well our predictions correlated with NMR order parameters to reflect motion. Our method had not been set up to address any of the tasks in those 4 case studies. Therefore, we expect that this method will assist in many attempts at inferring aspects of function.     

Efficient prediction of nucleic acid binding function from low-resolution protein structures

Structural genomics projects as well as ab initio protein structure prediction methods provide structures of proteins with no sequence or fold similarity to proteins with known functions. These are often low-resolution structures that may only include the positions of C alpha atoms. We present a fast and efficient method to predict DNA-binding proteins from just the amino acid sequences and low-resolution, C alpha-only protein models. The method uses the relative proportions of certain amino acids in the protein sequence, the asymmetry of the spatial distribution of certain other amino acids as well as the dipole moment of the molecule. These quantities are used in a linear formula, with coefficients derived from logistic regression performed on a training set, and DNA-binding is predicted based on whether the result is above a certain threshold. We show that the method is insensitive to errors in the atomic coordinates and provides correct predictions even on inaccurate protein models. We demonstrate that the method is capable of predicting proteins with novel binding site motifs and structures solved in an unbound state. The accuracy of our method is close to another, published method that uses all-atom structures, time-consuming calculations and information on conserved residues.     

Predicting disulfide connectivity from protein sequence using multiple sequence feature vectors and secondary structure

MOTIVATION: Disulfide bonds are primary covalent crosslinks between two cysteine residues in proteins that play critical roles in stabilizing the protein structures and are commonly found in extracy-toplasmatic or secreted proteins. In protein folding prediction, the localization of disulfide bonds can greatly reduce the search in conformational space. Therefore, there is a great need to develop computational methods capable of accurately predicting disulfide connectivity patterns in proteins that could have potentially important applications. RESULTS: We have developed a novel method to predict disulfide connectivity patterns from protein primary sequence, using a support vector regression (SVR) approach based on multiple sequence feature vectors and predicted secondary structure by the PSIPRED program. The results indicate that our method could achieve a prediction accuracy of 74.4% and 77.9%, respectively, when averaged on proteins with two to five disulfide bridges using 4-fold cross-validation, measured on the protein and cysteine pair on a well-defined non-homologous dataset. We assessed the effects of different sequence encoding schemes on the prediction performance of disulfide connectivity. It has been shown that the sequence encoding scheme based on multiple sequence feature vectors coupled with predicted secondary structure can significantly improve the prediction accuracy, thus enabling our method to outperform most of other currently available predictors. Our work provides a complementary approach to the current algorithms that should be useful in computationally assigning disulfide connectivity patterns and helps in the annotation of protein sequences generated by large-scale whole-genome projects. AVAILABILITY: The prediction web server and Supplementary Material are accessible at http://foo.maths.uq.edu.au/~huber/disulfide     

Statistical and molecular dynamics studies of buried waters in globular proteins

Buried solvent molecules are common in the core of globular proteins and contribute to structural stability. Folding necessitates the burial of polar backbone atoms in the protein core, whose hydrogen-bonding capacities should be satisfied on average. Whereas the residues in alpha-helices and beta-sheets form systematic main-chain hydrogen bonds, the residues in turns, coils and loops often contain polar atoms that fail to form intramolecular hydrogen bonds. The statistical analysis of 842 high resolution protein structures shows that well-resolved, internal water molecules preferentially reside near residues without alpha-helical and beta-sheet secondary structures. These buried waters most often form primary hydrogen bonds to main-chain atoms not involved in intramolecular hydrogen bonds, providing strong evidence that hydrating main-chain atoms is a key structural role of buried water molecules. Additionally, the average B-factor of protein atoms hydrogen-bonded to waters is smaller than that of protein atoms forming intramolecular hydrogen bonds, and the average B-factor of water molecules involved in primary hydrogen bonds with main-chain atoms is smaller than the average B-factor of water molecules involved in secondary hydrogen bonds to protein atoms that form concurrent intramolecular hydrogen bonds. To study the structural coupling between internal waters and buried polar atoms in detail we simulated the dynamics of wild-type FKBP12, in which a buried water, Wat137, forms one side-chain and multiple main-chain hydrogen bonds. We mutated E60, whose side-chain hydrogen bonds with Wat137, to Q, N, S or A, to modulate the multiplicity and geometry of hydrogen bonds to the water. Mutating E60 to a residue that is unable to form a hydrogen bond with Wat137 results in reorientation of the water molecule and leads to a structural readjustment of residues that are both near and distant to the water. We predict that the E60A mutation will result in a significantly reduced affinity of FKBP12 for its ligand FK506. The propensity of internal waters to hydrogen bond to buried polar atoms suggests that ordered water molecules may constitute fundamental structural components of proteins, particularly in regions where alpha-helical or beta-sheet secondary structure is not present.     

Prediction of protein accessible surface areas by support vector regression

A novel support vector regression (SVR) approach is proposed to predict protein accessible surface areas (ASAs) from their primary structures. In this work, we predict the real values of ASA in squared angstroms for residues instead of relative solvent accessibility. Based on protein residues, the mean and median absolute errors are 26.0 A(2) and 18.87 A(2), respectively. The correlation coefficient between the predicted and observed ASAs is 0.66. Cysteine is the best predicted amino acid (mean absolute error is 13.8 A(2) and median absolute error is 8.37 A(2)), while arginine is the least predicted amino acid (mean absolute error is 42.7 A(2) and median absolute error is 36.31 A(2)). Our work suggests that the SVR approach can be directly applied to the ASA prediction where data preclassification has been used.     

Predicting and annotating catalytic residues: an information theoretic approach.

We introduce a computational method to predict and annotate the catalytic residues of a protein using only its sequence information, so that we describe both the residues' sequence locations (prediction) and their specific biochemical roles in the catalyzed reaction (annotation). While knowing the chemistry of an enzyme's catalytic residues is essential to understanding its function, the challenges of prediction and annotation have remained difficult, especially when only the enzyme's sequence and no homologous structures are available. Our sequence-based approach follows the guiding principle that catalytic residues performing the same biochemical function should have similar chemical environments; it detects specific conservation patterns near in sequence to known catalytic residues and accordingly constrains what combination of amino acids can be present near a predicted catalytic residue. We associate with each catalytic residue a short sequence profile and define a Kullback-Leibler (KL) distance measure between these profiles, which, as we show, effectively captures even subtle biochemical variations. We apply the method to the class of glycohydrolase enzymes. This class includes proteins from 96 families with very different sequences and folds, many of which perform important functions. In a cross-validation test, our approach correctly predicts the location of the enzymes' catalytic residues with a sensitivity of 80% at a specificity of 99.4%, and in a separate cross-validation we also correctly annotate the biochemical role of 80% of the catalytic residues. Our results compare favorably to existing methods. Moreover, our method is more broadly applicable because it relies on sequence and not structure information; it may, furthermore, be used in conjunction with structure-based methods.     

Deriving protein dynamical properties from weighted protein contact number

It has recently been shown that in proteins the atomic mean-square displacement (or B-factor) can be related to the number of the neighboring atoms (or protein contact number), and that this relationship allows one to compute the B-factor profiles directly from protein contact number. This method, referred to as the protein contact model, is appealing, since it requires neither trajectory integration nor matrix diagonalization. As a result, the protein contact model can be applied to very large proteins and can be implemented as a high-throughput computational tool to compute atomic fluctuations in proteins. Here, we show that this relationship can be further refined to that between the atomic mean-square displacement and the weighted protein contact-number, the weight being the square of the reciprocal distance between the contacting pair. In addition, we show that this relationship can be utilized to compute the cross-correlation of atomic motion (the B-factor is essentially the auto-correlation of atomic motion). For a nonhomologous dataset comprising 972 high-resolution X-ray protein structures (resolution <2.0 A and sequence identity <25%), the mean correlation coefficient between the X-ray and computed B-factors based on the weighted protein contact-number model is 0.61, which is better than those of the original contact-number model (0.51) and other methods. We also show that the computed correlation maps based on the weighted contact-number model are globally similar to those computed through normal model analysis for some selected cases. Our results underscore the relationship between protein dynamics and protein packing. We believe that our method will be useful in the study of the protein structure-dynamics relationship.     

Towards predicting Ca2+‐binding sites with different coordination numbers in proteins with atomic resolution

Ca²⁺‐binding sites in proteins exhibit a wide range of polygonal geometries that directly relate to an equally‐diverse set of biological functions. Although the highly‐conserved EF‐Hand motif has been studied extensively, non‐EF‐Hand sites exhibit much more structural diversity which has inhibited efforts to determine the precise location of Ca²⁺‐binding sites, especially for sites with few coordinating ligands. Previously, we established an algorithm capable of predicting Ca²⁺‐binding sites using graph theory to identify oxygen clusters comprised of four atoms lying on a sphere of specified radius, the center of which was the predicted calcium position. Here we describe a new algorithm, MUG (MUltiple Geometries), which predicts Ca²⁺‐binding sites in proteins with atomic resolution. After first identifying all the possible oxygen clusters by finding maximal cliques, a calcium center (CC) for each cluster, corresponding to the potential Ca²⁺ position, is located to maximally regularize the structure of the (cluster, CC) pair. The structure is then inspected by geometric filters. An unqualified (cluster, CC) pair is further handled by recursively removing oxygen atoms and relocating the CC until its structure is either qualified or contains fewer than four ligand atoms. Ligand coordination is then determined for qualified structures. This algorithm, which predicts both Ca²⁺ positions and ligand groups, has been shown to successfully predict over 90% of the documented Ca²⁺‐binding sites in three datasets of highly‐diversified protein structures with 0.22 to 0.49 Å accuracy. All multiple‐binding sites (i.e. sites with a single ligand atom associated with multiple calcium ions) were predicted, as were half of the low‐coordination sites (i.e. sites with less than four protein ligand atoms) and 14/16 cofactor‐coordinating sites. Additionally, this algorithm has the flexibility to incorporate surface water molecules and protein cofactors to further improve the prediction for low‐coordination and cofactor‐coordinating Ca²⁺‐binding sites. Proteins 2009. © 2008 Wiley‐Liss, Inc.     

Using Support Vector Machine Combined with Post-processing Procedure to Improve Prediction of Interface Residues in Transient Complexes

Reliable prediction of interface residues in transient complexes remains challenging, yet is highly desirable for the design of new drugs. The existing computational methods mainly rely on evolutionary information to identify these key residues, but evolutionary information may not be effective for the interface residues in all types of transient complexes, such as antigen–antibody complexes. Herein we combined B-factor with sequence profile and accessible surface area to predict these important residues using support vector machine (SVM). Furthermore, a post-processing method was developed to reduce the number of false positives recognized by SVM. The prediction results show that B-factor is an effective indicator for the interface residues in antigen–antibody complexes as well as those in other types of transient complexes. In addition, we found that the post-processing procedure made an important contribution to further improve the prediction performance. Consequently, the proposed approach could provide new insight into accurately predicting interface residues in different types of transient complexes.     

Interactions in native binding sites cause a large change in protein dynamics

Cellular functions are regulated by molecules that interact with proteins and alter their activities. To enable such control, protein activity, and therefore protein conformational distributions, must be susceptible to alteration by molecular interactions at functional sites. Here we investigate whether interactions at functional sites cause a large change in the protein conformational distribution. We apply a computational method, called dynamics perturbation analysis (DPA), to identify sites at which interactions have a large allosteric potential D(x), which is the Kullback-Leibler divergence between protein conformational distributions with and without an interaction. In DPA, a protein is decorated with surface points that interact with neighboring protein atoms, and D(x) is calculated for each of the points in a coarse-grained model of protein vibrations. We use DPA to examine hundreds of protein structures from a standard small-molecule docking test set, and find that ligand-binding sites have elevated values of D(x): for 95% of proteins, the probability of randomly obtaining values as high as those in the binding site is 10(-3) or smaller. We then use DPA to develop a computational method to predict functional sites in proteins, and find that the method accurately predicts ligand-binding-site residues for proteins in the test set. The performance of this method compares favorably with that of a cleft analysis method. The results confirm that interactions at small-molecule binding sites cause a large change in the protein conformational distribution, and motivate using DPA for large-scale prediction of functional sites in proteins. They also suggest that natural selection favors proteins whose activities are capable of being regulated by molecular interactions.     

Prediction of helix-helix contacts and interacting helices in polytopic membrane proteins using neural networks

Despite rapidly increasing numbers of available 3D structures, membrane proteins still account for less than 1% of all structures in the Protein Data Bank. Recent high-resolution structures indicate a clearly broader structural diversity of membrane proteins than initially anticipated, motivating the development of reliable structure prediction methods specifically tailored for this class of molecules. One important prediction target capturing all major aspects of a protein's 3D structure is its contact map. Our analysis shows that computational methods trained to predict residue contacts in globular proteins perform poorly when applied to membrane proteins. We have recently published a method to identify interacting alpha-helices in membrane proteins based on the analysis of coevolving residues in predicted transmembrane regions. Here, we present a substantially improved algorithm for the same problem, which uses a newly developed neural network approach to predict helix-helix contacts. In addition to the input features commonly used for contact prediction of soluble proteins, such as windowed residue profiles and residue distance in the sequence, our network also incorporates features that apply to membrane proteins only, such as residue position within the transmembrane segment and its orientation toward the lipophilic environment. The obtained neural network can predict contacts between residues in transmembrane segments with nearly 26% accuracy. It is therefore the first published contact predictor developed specifically for membrane proteins performing with equal accuracy to state-of-the-art contact predictors available for soluble proteins. The predicted helix-helix contacts were employed in a second step to identify interacting helices. For our dataset consisting of 62 membrane proteins of solved structure, we gained an accuracy of 78.1%. Because the reliable prediction of helix interaction patterns is an important step in the classification and prediction of membrane protein folds, our method will be a helpful tool in compiling a structural census of membrane proteins.     

ProMate: a structure based prediction program to identify the location of protein-protein binding sites

Is the whole protein surface available for interaction with other proteins, or are specific sites pre-assigned according to their biophysical and structural character? And if so, is it possible to predict the location of the binding site from the surface properties? These questions are answered quantitatively by probing the surfaces of proteins using spheres of radius of 10 A on a database (DB) of 57 unique, non-homologous proteins involved in heteromeric, transient protein-protein interactions for which the structures of both the unbound and bound states were determined. In structural terms, we found the binding site to have a preference for beta-sheets and for relatively long non-structured chains, but not for alpha-helices. Chemically, aromatic side-chains show a clear preference for binding sites. While the hydrophobic and polar content of the interface is similar to the rest of the surface, hydrophobic and polar residues tend to cluster in interfaces. In the crystal, the binding site has more bound water molecules surrounding it, and a lower B-factor already in the unbound protein. The same biophysical properties were found to hold for the unbound and bound DBs. All the significant interface properties were combined into ProMate, an interface prediction program. This was followed by an optimization step to choose the best combination of properties, as many of them are correlated. During optimization and prediction, the tested proteins were not used for data collection, to avoid over-fitting. The prediction algorithm is fully automated, and is used to predict the location of potential binding sites on unbound proteins with known structures. The algorithm is able to successfully predict the location of the interface for about 70% of the proteins. The success rate of the predictor was equal whether applied on the unbound DB or on the disjoint bound DB. A prediction is assumed correct if over half of the predicted continuous interface patch is indeed interface. The ability to predict the location of protein-protein interfaces has far reaching implications both towards our understanding of specificity and kinetics of binding, as well as in assisting in the analysis of the proteome.     

Protein contacts, inter-residue interactions and side-chain modelling

Three-dimensional structures of proteins are the support of their biological functions. Their folds are stabilized by contacts between residues. Inner protein contacts are generally described through direct atomic contacts, i.e. interactions between side-chain atoms, while contact prediction methods mainly used inter-Calpha distances. In this paper, we have analyzed the protein contacts on a recent high quality non-redundant databank using different criteria. First, we have studied the average number of contacts depending on the distance threshold to define a contact. Preferential contacts between types of amino acids have been highlighted. Detailed analyses have been done concerning the proximity of contacts in the sequence, the size of the proteins and fold classes. The strongest differences have been extracted, highlighting important residues. Then, we studied the influence of five different side-chain conformation prediction methods (SCWRL, IRECS, SCAP, SCATD and SCCOMP) on the distribution of contacts. The prediction rates of these different methods are quite similar. However, using a distance criterion between side chains, the results are quite different, e.g. SCAP predicts 50% more contacts than observed, unlike other methods that predict fewer contacts than observed. Contacts deduced are quite distinct from one method to another with at most 75% contacts in common. Moreover, distributions of amino acid preferential contacts present unexpected behaviours distinct from previously observed in the X-ray structures, especially at the surface of proteins. For instance, the interactions involving Tryptophan greatly decrease.     

Effective scoring function for protein sequence design

We have developed an effective scoring function for protein design. The atomic solvation parameters, together with the weights of energy terms, were optimized so that residues corresponding to the native sequence were predicted with low energy in the training set of 28 protein structures. The solvation energy of non-hydrogen-bonded hydrophilic atoms was considered separately and expressed in a nonlinear way. As a result, our scoring function predicted native residues as the most favorable in 59% of the total positions in 28 proteins. We then tested the scoring function by comparing the predicted stability changes for 103 T4 lysozyme mutants with the experimental values. The correlation coefficients were 0.77 for surface mutations and 0.71 for all mutations. Finally, the scoring function combined with Monte Carlo simulation was used to predict favorable sequences on a fixed backbone. The designed sequences were similar to the natural sequences of the family to which the template structure belonged. The profile of the designed sequences was helpful for identification of remote homologues of the native sequence.     

Building proteins from C alpha coordinates using the dihedral probability grid Monte Carlo method.

Dihedral probability grid Monte Carlo (DPG-MC) is a general-purpose method of conformational sampling that can be applied to many problems in peptide and protein modeling. Here we present the DPG-MC method and apply it to predicting complete protein structures from C alpha coordinates. This is useful in such endeavors as homology modeling, protein structure prediction from lattice simulations, or fitting protein structures to X-ray crystallographic data. It also serves as an example of how DPG-MC can be applied to systems with geometric constraints. The conformational propensities for individual residues are used to guide conformational searches as the protein is built from the amino-terminus to the carboxyl-terminus. Results for a number of proteins show that both the backbone and side chain can be accurately modeled using DPG-MC. Backbone atoms are generally predicted with RMS errors of about 0.5 A (compared to X-ray crystal structure coordinates) and all atoms are predicted to an RMS error of 1.7 A or better.     

Role of evolutionary information in prediction of aromatic-backbone NH interactions in proteins

In this study, an attempt has been made to develop a neural network-based method for predicting segments in proteins containing aromatic-backbone NH (Ar-NH) interactions using multiple sequence alignment. We have analyzed 3121 segments seven residues long containing Ar-NH interactions, extracted from 2298 non-redundant protein structures where no two proteins have more than 25% sequence identity. Two consecutive feed-forward neural networks with a single hidden layer have been trained with standard back-propagation as learning algorithm. The performance of the method improves from 0.12 to 0.15 in terms of Matthews correlation coefficient (MCC) value when evolutionary information (multiple alignment obtained from PSI-BLAST) is used as input instead of a single sequence. The performance of the method further improves from MCC 0.15 to 0.20 when secondary structure information predicted by PSIPRED is incorporated in the prediction. The final network yields an overall prediction accuracy of 70.1% and an MCC of 0.20 when tested by five-fold cross-validation. Overall the performance is 15.2% higher than the random prediction. The method consists of two neural networks: (i) a sequence-to-structure network which predicts the aromatic residues involved in Ar-NH interaction from multiple alignment of protein sequences and (ii) a structure-to structure network where the input consists of the output obtained from the first network and predicted secondary structure. Further, the actual position of the donor residue within the 'potential' predicted fragment has been predicted using a separate sequence-to-structure neural network. Based on the present study, a server Ar_NHPred has been developed which predicts Ar-NH interaction in a given amino acid sequence. The web server Ar_NHPred is available at and (mirror site).     

Prediction of structures of zinc‐binding proteins through explicit modeling of metal coordination geometry

Metal ions play an essential role in stabilizing protein structures and contributing to protein function. Ions such as zinc have well‐defined coordination geometries, but it has not been easy to take advantage of this knowledge in protein structure prediction efforts. Here, we present a computational method to predict structures of zinc‐binding proteins given knowledge of the positions of zinc‐coordinating residues in the amino acid sequence. The method takes advantage of the “atom‐tree” representation of molecular systems and modular architecture of the Rosetta3 software suite to incorporate explicit metal ion coordination geometry into previously developed de novo prediction and loop modeling protocols. Zinc cofactors are tethered to their interacting residues based on coordination geometries observed in natural zinc‐binding proteins. The incorporation of explicit zinc atoms and their coordination geometry in both de novo structure prediction and loop modeling significantly improves sampling near the native conformation. The method can be readily extended to predict protein structures bound to other metal and/or small chemical cofactors with well‐defined coordination or ligation geometry.