Consensus scoring for enriching near-native structures from protein-protein docking decoys

The identification of near native protein-protein complexes among a set of decoys remains highly challenging. A strategy for improving the success rate of near native detection is to enrich near native docking decoys in a small number of top ranked decoys. Recently, we found that a combination of three scoring functions (energy, conservation, and interface propensity) can predict the location of binding interface regions with reasonable accuracy. Here, these three scoring functions are modified and combined into a consensus scoring function called ENDES for enriching near native docking decoys. We found that all individual scores result in enrichment for the majority of 28 targets in ZDOCK2.3 decoy set and the 22 targets in Benchmark 2.0. Among the three scores, the interface propensity score yields the highest enrichment in both sets of protein complexes. When these scores are combined into the ENDES consensus score, a significant increase in enrichment of near-native structures is found. For example, when 2000 dock decoys are reduced to 200 decoys by ENDES, the fraction of near-native structures in docking decoys increases by a factor of about six in average. ENDES was implemented into a computer program that is available for download at http://sparks.informatics.iupui.edu.     

Discrimination of near-native structures in protein-protein docking by testing the stability of local minima

Fast Fourier transform (FFT) correlation methods of protein-protein docking, combined with the clustering of low energy conformations, can find a number of local minima on the energy surface. For most complexes, the locations of the near-native structures can be constrained to the 30 largest clusters, each surrounding a local minimum. However, no reliable further discrimination can be obtained by energy measures because the differences in the energy levels between the minima are comparable with the errors in the energy evaluation. In fact, no current scoring function accounts for the entropic contributions that relate to the width rather than the depth of the minima. Since structures at narrow minima loose more entropy, some of the nonnative states can be detected by determining whether or not a local minimum is surrounded by a broad region of attraction on the energy surface. The analysis is based on starting Monte Carlo Minimization (MCM) runs from random points around each minimum, and observing whether a certain fraction of trajectories converge to a small region within the cluster. The cluster is considered stable if such a strong attractor exists, has at least 10 convergent trajectories, is relatively close to the original cluster center, and contains a low energy structure. We studied the stability of clusters for enzyme-inhibitor and antibody-antigen complexes in the Protein Docking Benchmark. The analysis yields three main results. First, all clusters that are close to the native structure are stable. Second, restricting considerations to stable clusters eliminates around half of the false positives, that is, solutions that are low in energy but far from the native structure of the complex. Third, dividing the conformational space into clusters and determining the stability of each cluster, the combined approach is less dependent on a priori information than exploring the potential conformational space by Monte Carlo minimizations.     

Identification of near-native structures by clustering protein docking conformations

Most state-of-the-art protein-protein docking algorithms use the Fast Fourier Transform (FFT) technique to sample the six-dimensional translational and rotational space. Scoring functions including shape complementarity, electrostatics, and desolvation are usually exploited in ranking the docking conformations. While these rigid-body docking methods provide good performance in bound docking, using unbound structures as input frequently leads to a high number of false positive hits. For the purpose of better selecting correct docking conformations, we structurally cluster the docking decoys generated by four widely-used FFT-based protein-protein docking methods. In all cases, the selection based on cluster size outperforms the ranking based on the inherent scoring function. If we cluster decoys from different servers together, only marginal improvement is obtained in comparison with clustering decoys from the best individual server. A collection of multiple decoy sets of comparable quality will be the key to improve the clustering result from meta-docking servers.     

Optimal clustering for detecting near-native conformations in protein docking

Clustering is one of the most powerful tools in computational biology. The conventional wisdom is that events that occur in clusters are probably not random. In protein docking, the underlying principle is that clustering occurs because long-range electrostatic and/or desolvation forces steer the proteins to a low free-energy attractor at the binding region. Something similar occurs in the docking of small molecules, although in this case shorter-range van der Waals forces play a more critical role. Based on the above, we have developed two different clustering strategies to predict docked conformations based on the clustering properties of a uniform sampling of low free-energy protein-protein and protein-small molecule complexes. We report on significant improvements in the automated prediction and discrimination of docked conformations by using the cluster size and consensus as a ranking criterion. We show that the success of clustering depends on identifying the appropriate clustering radius of the system. The clustering radius for protein-protein complexes is consistent with the range of the electrostatics and desolvation free energies (i.e., between 4 and 9 Angstroms); for protein-small molecule docking, the radius is set by van der Waals interactions (i.e., at approximately 2 Angstroms). Without any a priori information, a simple analysis of the histogram of distance separations between the set of docked conformations can evaluate the clustering properties of the data set. Clustering is observed when the histogram is bimodal. Data clustering is optimal if one chooses the clustering radius to be the minimum after the first peak of the bimodal distribution. We show that using this optimal radius further improves the discrimination of near-native complex structures.     

Accelerating and focusing protein-protein docking correlations using multi-dimensional rotational FFT generating functions

MOTIVATION: Predicting how proteins interact at the molecular level is a computationally intensive task. Many protein docking algorithms begin by using fast Fourier transform (FFT) correlation techniques to find putative rigid body docking orientations. Most such approaches use 3D Cartesian grids and are therefore limited to computing three dimensional (3D) translational correlations. However, translational FFTs can speed up the calculation in only three of the six rigid body degrees of freedom, and they cannot easily incorporate prior knowledge about a complex to focus and hence further accelerate the calculation. Furthemore, several groups have developed multi-term interaction potentials and others use multi-copy approaches to simulate protein flexibility, which both add to the computational cost of FFT-based docking algorithms. Hence there is a need to develop more powerful and more versatile FFT docking techniques. RESULTS: This article presents a closed-form 6D spherical polar Fourier correlation expression from which arbitrary multi-dimensional multi-property multi-resolution FFT correlations may be generated. The approach is demonstrated by calculating 1D, 3D and 5D rotational correlations of 3D shape and electrostatic expansions up to polynomial order L=30 on a 2 GB personal computer. As expected, 3D correlations are found to be considerably faster than 1D correlations but, surprisingly, 5D correlations are often slower than 3D correlations. Nonetheless, we show that 5D correlations will be advantageous when calculating multi-term knowledge-based interaction potentials. When docking the 84 complexes of the Protein Docking Benchmark, blind 3D shape plus electrostatic correlations take around 30 minutes on a contemporary personal computer and find acceptable solutions within the top 20 in 16 cases. Applying a simple angular constraint to focus the calculation around the receptor binding site produces acceptable solutions within the top 20 in 28 cases. Further constraining the search to the ligand binding site gives up to 48 solutions within the top 20, with calculation times of just a few minutes per complex. Hence the approach described provides a practical and fast tool for rigid body protein-protein docking, especially when prior knowledge about one or both binding sites is available.     

Improved side-chain modeling for protein-protein docking

Success in high-resolution protein-protein docking requires accurate modeling of side-chain conformations at the interface. Most current methods either leave side chains fixed in the conformations observed in the unbound protein structures or allow the side chains to sample a set of discrete rotamer conformations. Here we describe a rapid and efficient method for sampling off-rotamer side-chain conformations by torsion space minimization during protein-protein docking starting from discrete rotamer libraries supplemented with side-chain conformations taken from the unbound structures, and show that the new method improves side-chain modeling and increases the energetic discrimination between good and bad models. Analysis of the distribution of side-chain interaction energies within and between the two protein partners shows that the new method leads to more native-like distributions of interaction energies and that the neglect of side-chain entropy produces a small but measurable increase in the number of residues whose interaction energy cannot compensate for the entropic cost of side-chain freezing at the interface. The power of the method is highlighted by a number of predictions of unprecedented accuracy in the recent CAPRI (Critical Assessment of PRedicted Interactions) blind test of protein-protein docking methods.     

Hydrophobic complementarity in protein-protein docking

Formation of hydrophobic contacts across a newly formed interface is energetically favorable. Based on this observation we developed a geometric-hydrophobic docking algorithm that estimates quantitatively the hydrophobic complementarity at protein-protein interfaces. Each molecule to be docked is represented as a grid of complex numbers, storing information regarding the shape of the molecule in the real part and information regarding the hydropathy of the surface in the imaginary part. The grid representations are correlated using fast Fourier transformations. The algorithm is used to compare the extent of hydrophobic complementarity in oligomers (represented by D2 tetramers) and in hetero-dimers of soluble proteins (complexes). We also test the implication of hydrophobic complementarity in distinguishing correct from false docking solutions. We find that hydrophobic complementarity at the interface exists in oligomers and in complexes, and in both groups the extent of such complementarity depends on the size of the interface. Thus, the non-polar portions of large interfaces are more often juxtaposed than non-polar portions of small interfaces. Next we find that hydrophobic complementarity helps to point out correct docking solutions. In oligomers it significantly improves the ranks of nearly correct reassembled and modeled tetramers. Combining geometric, electrostatic and hydrophobic complementarity for complexes gives excellent results, ranking a nearly correct solution < 10 for 5 of 23 tested systems, < 100 for 8 systems and < 1000 for 19 systems.     

Inferring protein-protein interacting sites using residue conservation and evolutionary information

This paper proposes a novel method using protein residue conservation and evolution information, i.e., spatial sequence profile, sequence information entropy and evolution rate, to infer protein binding sites. Some predictors based on support vector machines (SVMs) algorithm are constructed to predict the role of surface residues in protein-protein interface. By combining protein residue characters, the prediction performance can be improved obviously. We then made use of the predicted labels of neighbor residues to improve the performance of the predictors. The efficiency and the effectiveness of our proposed approach are verified by its better prediction performance based on a non-redundant data set of heterodimers.     

New benchmark metrics for protein-protein docking methods

With the development of many computational methods that predict the structural models of protein-protein complexes, there is a pressing need to benchmark their performance. As was the case for protein monomers, assessing the quality of models of protein complexes is not straightforward. An effective scoring scheme should be able to detect substructure similarity and estimate its statistical significance. Here, we focus on characterizing the similarity of the interfaces of the complex and introduce two scoring functions. The first, the interfacial Template Modeling score (iTM-score), measures the geometric distance between the interfaces, while the second, the Interface Similarity score (IS-score), evaluates their residue-residue contact similarity in addition to their geometric similarity. We first demonstrate that the IS-score is more suitable for assessing docking models than the iTM-score. The IS-score is then validated in a large-scale benchmark test on 1562 dimeric complexes. Finally, the scoring function is applied to evaluate docking models submitted to the Critical Assessment of Prediction of Interactions (CAPRI) experiments. While the results according to the new scoring scheme are generally consistent with the original CAPRI assessment, the IS-score identifies models whose significance was previously underestimated.     

Accounting for loop flexibility during protein-protein docking

Although reliable docking can now be achieved for systems that do not undergo important induced conformational change upon association, the presence of flexible surface loops, which must adapt to the steric and electrostatic properties of a partner, generally presents a major obstacle. We report here the first docking method that allows large loop movements during a systematic exploration of the possible arrangements of the two partners in terms of position and rotation. Our strategy consists in taking into account an ensemble of possible loop conformations by a multi-copy representation within a reduced protein model. The docking process starts from regularly distributed positions and orientations of the ligand around the whole receptor. Each starting configuration is submitted to energy minimization during which the best-fitting loop conformation is selected based on the mean-field theory. Trials were carried out on proteins with significant differences in the main-chain conformation of the binding loop between isolated form and complexed form, which were docked to their partner considered in their bound form. The method is able to predict complexes very close to the crystal complex both in terms of relative position of the two partners and of the geometry of the flexible loop. We also show that introducing loop flexibility on the isolated protein form during systematic docking largely improves the predictions of relative position of the partners in comparison with rigid-body docking.     

Complex-type-dependent scoring functions in protein-protein docking

A major challenge in the field of protein-protein docking is to discriminate between the many wrong and few near-native conformations, i.e. scoring. Here, we introduce combinatorial complex-type-dependent scoring functions for different types of protein-protein complexes, protease/inhibitor, antibody/antigen, enzyme/inhibitor and others. The scoring functions incorporate both physical and knowledge-based potentials, i.e. atomic contact energy (ACE), the residue pair potential (RP), electrostatic and van der Waals' interactions. For different type complexes, the weights of the scoring functions were optimized by the multiple linear regression method, in which only top 300 structures with ligand root mean square deviation (L_RMSD) less than 20 A from the bound (co-crystallized) docking of 57 complexes were used to construct a training set. We employed the bound docking studies to examine the quality of the scoring function, and also extend to the unbound (separately crystallized) docking studies and extra 8 protein-protein complexes. In bound docking of the 57 cases, the first hits of protease/inhibitor cases are all ranked in the top 5. For the cases of antibody/antigen, enzyme/inhibitor and others, there are 17/19, 5/6 and 13/15 cases with the first hits ranked in the top 10, respectively. In unbound docking studies, the first hits of 9/17 protease/inhibitor, 6/19 antibody/antigen, 1/6 enzyme/inhibitor and 6/15 others' complexes are ranked in the top 10. Additionally, for the extra 8 cases, the first hits of the two protease/inhibitor cases are ranked in the top for the bound and unbound test. For the two enzyme/inhibitor cases, the first hits are ranked 1st for bound test, and the 119th and 17th for the unbound test. For the others, the ranks of the first hits are the 1st for the bound test and the 12th for the 1WQ1 unbound test. To some extent, the results validated our divide-and-conquer strategy in the docking study, which might hopefully shed light on the prediction of protein-protein interactions.     

Density-based score for selecting near-native atomic models of unknown structures

We present a low-resolution density-based scoring scheme for selecting high-quality models from a large pool of lesser quality models. We use pre-configured decoy data sets that contain large numbers of models with different degrees of correctness to evaluate the performance of the strategy. We find that the scoring scheme consistently identifies one of the highest quality models for a wide variety of target structures, resolution ranges, and noise models. Tests with experimental data yield similar results.     

Designing coarse grained-and atom based-potentials for protein-protein docking

Background  

Protein-protein docking is a challenging computational problem in functional genomics, particularly when one or both proteins undergo conformational change(s) upon binding. The major challenge is to define a scoring function soft enough to tolerate these changes and specific enough to distinguish between near-native and "misdocked" conformations.     

Database driven test case generation for protein-protein docking

SUMMARY: We present a method for automatic test case generation for protein-protein docking. A consensus-type approach is proposed processing the whole PDB and classifying protein structures into complexes and unbound proteins by combining information from three different approaches (current PDB-at-a-glance classification, search of complexes by sequence identical unbound structures and chain naming). Out of this classification test cases are generated automatically. All calculations were run on the database. The information stored is available via a web interface. The user can choose several criteria for generating his own subset out of our test cases, e.g. for testing docking algorithms. AVAILABILITY: http://bibiserv.techfak.uni-bielefeld.de/agt-sdp/ CONTACT: fzoellne@techfak.uni-bielefeld.de.     

A collaborative filtering approach for protein-protein docking scoring functions

A protein-protein docking procedure traditionally consists in two successive tasks: a search algorithm generates a large number of candidate conformations mimicking the complex existing in vivo between two proteins, and a scoring function is used to rank them in order to extract a native-like one. We have already shown that using Voronoi constructions and a well chosen set of parameters, an accurate scoring function could be designed and optimized. However to be able to perform large-scale in silico exploration of the interactome, a near-native solution has to be found in the ten best-ranked solutions. This cannot yet be guaranteed by any of the existing scoring functions. In this work, we introduce a new procedure for conformation ranking. We previously developed a set of scoring functions where learning was performed using a genetic algorithm. These functions were used to assign a rank to each possible conformation. We now have a refined rank using different classifiers (decision trees, rules and support vector machines) in a collaborative filtering scheme. The scoring function newly obtained is evaluated using 10 fold cross-validation, and compared to the functions obtained using either genetic algorithms or collaborative filtering taken separately. This new approach was successfully applied to the CAPRI scoring ensembles. We show that for 10 targets out of 12, we are able to find a near-native conformation in the 10 best ranked solutions. Moreover, for 6 of them, the near-native conformation selected is of high accuracy. Finally, we show that this function dramatically enriches the 100 best-ranking conformations in near-native structures.     

A novel shape complementarity scoring function for protein-protein docking

Shape complementarity is the most basic ingredient of the scoring functions for protein-protein docking. Most grid-based docking algorithms use the total number of grid points at the binding interface to quantify shape complementarity. We have developed a novel Pairwise Shape Complementarity (PSC) function that is conceptually simple and rapid to compute. The favorable component of PSC is the total number of atom pairs between the receptor and the ligand within a distance cutoff. When applied to a benchmark of 49 test cases, PSC consistently ranks near-native structures higher and produces more near-native structures than the traditional grid-based function, and the improvement was seen across all prediction levels and in all categories of the benchmark. Without any post-processing or biological information about the binding site except the complementarity-determining region of antibodies, PSC predicts the complex structure correctly for 6 test cases, and ranks at least one near-native structure in the top 20 predictions for 18 test cases. Our docking program ZDOCK has been parallelized and the average computing time is 4 minutes using sixteen IBM SP3 processors. Both ZDOCK and the benchmark are freely available to academic users (http://zlab.bu.edu/~ rong/dock).     

Integrating atom-based and residue-based scoring functions for protein-protein docking

Most scoring functions for protein-protein docking algorithms are either atom-based or residue-based, with the former being able to produce higher quality structures and latter more tolerant to conformational changes upon binding. Earlier, we developed the ZRANK algorithm for reranking docking predictions, with a scoring function that contained only atom-based terms. Here we combine ZRANK's atom-based potentials with five residue-based potentials published by other labs, as well as an atom-based potential IFACE that we published after ZRANK. We simultaneously optimized the weights for selected combinations of terms in the scoring function, using decoys generated with the protein-protein docking algorithm ZDOCK. We performed rigorous cross validation of the combinations using 96 test cases from a docking benchmark. Judged by the integrative success rate of making 1000 predictions per complex, addition of IFACE and the best residue-based pair potential reduced the number of cases without a correct prediction by 38 and 27% relative to ZDOCK and ZRANK, respectively. Thus combination of residue-based and atom-based potentials into a scoring function can improve performance for protein-protein docking. The resulting scoring function is called IRAD (integration of residue- and atom-based potentials for docking) and is available at http://zlab.umassmed.edu.