Enabling structure-based drug discovery utilizing predicted models.

High-quality predicted structures enable structure-based approaches to an expanding number of drug discovery programs. We propose that by utilizing free energy perturbation (FEP), predicted structures can be confidently employed to achieve drug design goals. We use structure-based modeling of hERG inhibition to illustrate this value of FEP.

Structure-based prediction of ligand-protein interactions on a genome-wide scale.

We report a template-based method, LT-scanner, which scans the human proteome using protein structural alignment to identify proteins that are likely to bind ligands that are present in experimentally determined complexes. A scoring function that rapidly accounts for binding site similarities between the template and the proteins being scanned is a crucial feature of the method. The overall approach is first tested based on its ability to predict the residues on the surface of a protein that are likely to bind small-molecule ligands. The algorithm that we present, LBias, is shown to compare very favorably to existing algorithms for binding site residue prediction. LT-scanner's performance is evaluated based on its ability to identify known targets of Food and Drug Administration (FDA)-approved drugs and it too proves to be highly effective. The specificity of the scoring function that we use is demonstrated by the ability of LT-scanner to identify the known targets of FDA-approved kinase inhibitors based on templates involving other kinases. Combining sequence with structural information further improves LT-scanner performance. The approach we describe is extendable to the more general problem of identifying binding partners of known ligands even if they do not appear in a structurally determined complex, although this will require the integration of methods that combine protein structure and chemical compound databases.

Evaluating template-based and template-free protein-protein complex structure prediction.

We compared the performance of template-free (docking) and template-based methods for the prediction of protein-protein complex structures. We found similar performance for a template-based method based on threading (COTH) and another template-based method based on structural alignment (PRISM). The template-based methods showed similar performance to a docking method (ZDOCK) when the latter was allowed one prediction for each complex, but when the same number of predictions was allowed for each method, the docking approach outperformed template-based approaches. We identified strengths and weaknesses in each method. Template-based approaches were better able to handle complexes that involved conformational changes upon binding. Furthermore, the threading-based and docking methods were better than the structural-alignment-based method for enzyme-inhibitor complex prediction. Finally, we show that the near-native (correct) predictions were generally not shared by the various approaches, suggesting that integrating their results could be the superior strategy.

ZDOCK server: interactive docking prediction of protein-protein complexes and symmetric multimers.

SUMMARY: Protein-protein interactions are essential to cellular and immune function, and in many cases, because of the absence of an experimentally determined structure of the complex, these interactions must be modeled to obtain an understanding of their molecular basis. We present a user-friendly protein docking server, based on the rigid-body docking programs ZDOCK and M-ZDOCK, to predict structures of protein-protein complexes and symmetric multimers. With a goal of providing an accessible and intuitive interface, we provide options for users to guide the scoring and the selection of output models, in addition to dynamic visualization of input structures and output docking models. This server enables the research community to easily and quickly produce structural models of protein-protein complexes and symmetric multimers for their own analysis. AVAILABILITY: The ZDOCK server is freely available to all academic and non-profit users at: http://zdock.umassmed.edu. No registration is required.

Binding interface prediction by combining protein-protein docking results.

We developed a method called residue contact frequency (RCF), which uses the complex structures generated by the protein-protein docking algorithm ZDOCK to predict interface residues. Unlike interface prediction algorithms that are based on monomers alone, RCF is binding partner specific. We evaluated the performance of RCF using the area under the precision-recall (PR) curve (AUC) on a large protein docking Benchmark. RCF (AUC = 0.44) performed as well as meta-PPISP (AUC = 0.43), which is one of the best monomer-based interface prediction methods. In addition, we test a support vector machine (SVM) to combine RCF with meta-PPISP and another monomer-based interface prediction algorithm Evolutionary Trace to further improve the performance. We found that the SVM that combined RCF and meta-PPISP achieved the best performance (AUC = 0.47). We used RCF to predict the binding interfaces of proteins that can bind to multiple partners and RCF was able to correctly predict interface residues that are unique for the respective binding partners. Furthermore, we found that residues that contributed greatly to binding affinity (hotspot residues) had significantly higher RCF than other residues.

Blind prediction of interfacial water positions in CAPRI.

We report the first assessment of blind predictions of water positions at protein-protein interfaces, performed as part of the critical assessment of predicted interactions (CAPRI) community-wide experiment. Groups submitting docking predictions for the complex of the DNase domain of colicin E2 and Im2 immunity protein (CAPRI Target 47), were invited to predict the positions of interfacial water molecules using the method of their choice. The predictions-20 groups submitted a total of 195 models-were assessed by measuring the recall fraction of water-mediated protein contacts. Of the 176 high- or medium-quality docking models-a very good docking performance per se-only 44% had a recall fraction above 0.3, and a mere 6% above 0.5. The actual water positions were in general predicted to an accuracy level no better than 1.5 Å, and even in good models about half of the contacts represented false positives. This notwithstanding, three hotspot interface water positions were quite well predicted, and so was one of the water positions that is believed to stabilize the loop that confers specificity in these complexes. Overall the best interface water predictions was achieved by groups that also produced high-quality docking models, indicating that accurate modelling of the protein portion is a determinant factor. The use of established molecular mechanics force fields, coupled to sampling and optimization procedures also seemed to confer an advantage. Insights gained from this analysis should help improve the prediction of protein-water interactions and their role in stabilizing protein complexes.

Performance of ZDOCK in CAPRI rounds 20-26.

We report the performance of our approaches for protein-protein docking and interface analysis in CAPRI rounds 20-26. At the core of our pipeline was the ZDOCK program for rigid-body protein-protein docking. We then reranked the ZDOCK predictions using the ZRANK or IRAD scoring functions, pruned and analyzed energy landscapes using clustering, and analyzed the docking results using our interface prediction approach RCF. When possible, we used biological information from the literature to apply constraints to the search space during or after the ZDOCK runs. For approximately half of the standard docking challenges we made at least one prediction that was acceptable or better. For the scoring challenges we made acceptable or better predictions for all but one target. This indicates that our scoring functions are generally able to select the correct binding mode.

PrePCI: A structure- and chemical similarity-informed database of predicted protein compound interactions.

We describe the Predicting Protein-Compound Interactions (PrePCI) database which comprises over 5 billion predicted interactions between 6.8 million chemical compounds and 19,797 human proteins. PrePCI relies on a proteome-wide database of structural models based on both traditional modeling techniques and the AlphaFold Protein Structure Database. Sequence- and structural similarity-based metrics are established between template proteins, T, in the Protein Data Bank that bind compounds, C, and query proteins in the model database, Q. When the metrics exceed threshold values, it is assumed that C also binds to Q with a likelihood ratio (LR) derived from machine learning. If the relationship is based on structural similarity, the LR is based on a scoring function that measures the extent to which C is compatible with the binding site of Q as described in the LT-scanner algorithm. For every predicted complex derived in this way, chemical similarity based on the Tanimoto coefficient identifies other small molecules that may bind to Q. An overall LR for the binding of C to Q is obtained from Naive Bayesian statistics. The PrePCI database can be queried by entering a UniProt ID or gene name for a protein to obtain a list of compounds predicted to bind to it along with associated LRs. Alternatively, entering an identifier for the compound outputs a list of proteins it is predicted to bind. Specific applications of the database to lead discovery, elucidation of drug mechanism of action, and biological function annotation are described.

A machine learning approach for the prediction of protein surface loop flexibility.

Proteins often undergo conformational changes when binding to each other. A major fraction of backbone conformational changes involves motion on the protein surface, particularly in loops. Accounting for the motion of protein surface loops represents a challenge for protein-protein docking algorithms. A first step in addressing this challenge is to distinguish protein surface loops that are likely to undergo backbone conformational changes upon protein-protein binding (mobile loops) from those that are not (stationary loops). In this study, we developed a machine learning strategy based on support vector machines (SVMs). Our SVM uses three features of loop residues in the unbound protein structures-Ramachandran angles, crystallographic B-factors, and relative accessible surface area-to distinguish mobile loops from stationary ones. This method yields an average prediction accuracy of 75.3% compared with a random prediction accuracy of 50%, and an average of 0.79 area under the receiver operating characteristic (ROC) curve using cross-validation. Testing the method on an independent dataset, we obtained a prediction accuracy of 70.5%. Finally, we applied the method to 11 complexes that involve members from the Ras superfamily and achieved prediction accuracy of 92.8% for the Ras superfamily proteins and 74.4% for their binding partners.

Protein-protein docking benchmark version 4.0.

We updated our protein-protein docking benchmark to include complexes that became available since our previous release. As before, we only considered high-resolution complex structures that are nonredundant at the family-family pair level, for which the X-ray or NMR unbound structures of the constituent proteins are also available. Benchmark 4.0 adds 52 new complexes to the 124 cases of Benchmark 3.0, representing an increase of 42%. Thus, benchmark 4.0 provides 176 unbound-unbound cases that can be used for protein-protein docking method development and assessment. Seventeen of the newly added cases are enzyme-inhibitor complexes, and we found no new antigen-antibody complexes. Classifying the new cases according to expected difficulty for protein-protein docking algorithms gives 33 rigid body cases, 11 cases of medium difficulty, and 8 cases that are difficult. Benchmark 4.0 listings and processed structure files are publicly accessible at http://zlab.umassmed.edu/benchmark/.

Performance of ZDOCK and ZRANK in CAPRI rounds 13-19.

We report the performance of the ZDOCK and ZRANK algorithms in CAPRI rounds 13-19 and introduce a novel measure atom contact frequency (ACF). To compute ACF, we identify the residues that most often make contact with the binding partner in the complete set of ZDOCK predictions for each target. We used ACF to predict the interface of the proteins, which, in combination with the biological data available in the literature, is a valuable addition to our docking pipeline. Furthermore, we incorporated a straightforward and efficient clustering algorithm with two purposes: (1) to determine clusters of similar docking poses (corresponding to energy funnels) and (2) to remove redundancies from the final set of predictions. With these new developments, we achieved at least one acceptable prediction for targets 29 and 36, at least one medium-quality prediction for targets 41 and 42, and at least one high-quality prediction for targets 37 and 40; thus, we succeeded for six out of a total of 12 targets.

Protein-protein docking benchmark version 3.0.

We present version 3.0 of our publicly available protein-protein docking benchmark. This update includes 40 new test cases, representing a 48% increase from Benchmark 2.0. For all of the new cases, the crystal structures of both binding partners are available. As with Benchmark 2.0, Structural Classification of Proteins (Murzin et al., J Mol Biol 1995;247:536-540) was used to remove redundant test cases. The 124 unbound-unbound test cases in Benchmark 3.0 are classified into 88 rigid-body cases, 19 medium-difficulty cases, and 17 difficult cases, based on the degree of conformational change at the interface upon complex formation. In addition to providing the community with more test cases for evaluating docking methods, the expansion of Benchmark 3.0 will facilitate the development of new algorithms that require a large number of training examples. Benchmark 3.0 is available to the public at http://zlab.bu.edu/benchmark.

The performance of ZDOCK and ZRANK in rounds 6-11 of CAPRI.

We present an evaluation of our protein-protein docking approach using the ZDOCK and ZRANK algorithms, in combination with structural clustering and filtering, utilizing biological data in Rounds 6-11 of the CAPRI docking experiment. We achieved at least one prediction of acceptable accuracy for five of six targets submitted. In addition, two targets resulted in medium-accuracy predictions. In the new scoring portion of the CAPRI exercise, we were able to attain at least one acceptable prediction for the three targets submitted and achieved three medium-accuracy predictions for Target 26. Scoring was performed using ZRANK, a new algorithm for reranking initial-stage docking predictions using a weighted energy function and no structural refinement. Here we outline a practical and successful docking strategy, given limited prior biological knowledge of the complex to be predicted.

A hybrid method for protein-protein interface prediction.

The growing structural coverage of proteomes is making structural comparison a powerful tool for function annotation. Such template-based approaches are based on the observation that structural similarity is often sufficient to infer similar function. However, it seems clear that, in addition to structural similarity, the specific characteristics of a given protein should also be taken into account in predicting function. Here we describe PredUs 2.0, a method to predict regions on a protein surface likely to bind other proteins, that is, interfacial residues. PredUs 2.0 is based on the PredUs method that is entirely template-based and uses known binding sites in structurally similar proteins to predict interfacial residues. PredUs 2.0 uses a Bayesian approach to combine the template-based scoring of PredUs with a score that reflects the propensities of individual amino acids to be in interfaces. PredUs 2.0 includes a novel protein size dependent metric to determine the number of residues that should be reported as interfacial. PredUs 2.0 significantly outperforms PredUs as well as other published interface prediction methods.

Template-based prediction of protein function.

We discuss recent approaches for structure-based protein function annotation. We focus on template-based methods where the function of a query protein is deduced from that of a template for which both the structure and function are known. We describe the different ways of identifying a template. These are typically based on sequence analysis but new methods based on purely structural similarity are also being developed that allow function annotation based on structural relationships that cannot be recognized by sequence. The growing number of available structures of known function, improved homology modeling techniques and new developments in the use of structure allow template-based methods to be applied on a proteome-wide scale and in many different biological contexts. This progress significantly expands the range of applicability of structural information in function annotation to a level that previously was only achievable by sequence comparison.

Exploring angular distance in protein-protein docking algorithms.

We present a two-stage hybrid-resolution approach for rigid-body protein-protein docking. The first stage is carried out at low-resolution (15°) angular sampling. In the second stage, we sample promising regions from the first stage at a higher resolution of 6°. The hybrid-resolution approach produces the same results as a 6° uniform sampling docking run, but uses only 17% of the computational time. We also show that the angular distance can be used successfully in clustering and pruning algorithms, as well as the characterization of energy funnels. Traditionally the root-mean-square-distance is used in these algorithms, but the evaluation is computationally expensive as it depends on both the rotational and translational parameters of the docking solutions. In contrast, the angular distances only depend on the rotational parameters, which are generally fixed for all docking runs. Hence the angular distances can be pre-computed, and do not add computational time to the post-processing of rigid-body docking results.

Prediction of protein-protein binding free energies.

We present an energy function for predicting binding free energies of protein-protein complexes, using the three-dimensional structures of the complex and unbound proteins as input. Our function is a linear combination of nine terms and achieves a correlation coefficient of 0.63 with experimental measurements when tested on a benchmark of 144 complexes using leave-one-out cross validation. Although we systematically tested both atomic and residue-based scoring functions, the selected function is dominated by residue-based terms. Our function is stable for subsets of the benchmark stratified by experimental pH and extent of conformational change upon complex formation, with correlation coefficients ranging from 0.61 to 0.66.

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.

A structure-based benchmark for protein-protein binding affinity.

We have assembled a nonredundant set of 144 protein-protein complexes that have high-resolution structures available for both the complexes and their unbound components, and for which dissociation constants have been measured by biophysical methods. The set is diverse in terms of the biological functions it represents, with complexes that involve G-proteins and receptor extracellular domains, as well as antigen/antibody, enzyme/inhibitor, and enzyme/substrate complexes. It is also diverse in terms of the partners' affinity for each other, with K(d) ranging between 10(-5) and 10(-14) M. Nine pairs of entries represent closely related complexes that have a similar structure, but a very different affinity, each pair comprising a cognate and a noncognate assembly. The unbound structures of the component proteins being available, conformation changes can be assessed. They are significant in most of the complexes, and large movements or disorder-to-order transitions are frequently observed. The set may be used to benchmark biophysical models aiming to relate affinity to structure in protein-protein interactions, taking into account the reactants and the conformation changes that accompany the association reaction, instead of just the final product.

Community-wide assessment of protein-interface modeling suggests improvements to design methodology.

The CAPRI (Critical Assessment of Predicted Interactions) and CASP (Critical Assessment of protein Structure Prediction) experiments have demonstrated the power of community-wide tests of methodology in assessing the current state of the art and spurring progress in the very challenging areas of protein docking and structure prediction. We sought to bring the power of community-wide experiments to bear on a very challenging protein design problem that provides a complementary but equally fundamental test of current understanding of protein-binding thermodynamics. We have generated a number of designed protein-protein interfaces with very favorable computed binding energies but which do not appear to be formed in experiments, suggesting that there may be important physical chemistry missing in the energy calculations. A total of 28 research groups took up the challenge of determining what is missing: we provided structures of 87 designed complexes and 120 naturally occurring complexes and asked participants to identify energetic contributions and/or structural features that distinguish between the two sets. The community found that electrostatics and solvation terms partially distinguish the designs from the natural complexes, largely due to the nonpolar character of the designed interactions. Beyond this polarity difference, the community found that the designed binding surfaces were, on average, structurally less embedded in the designed monomers, suggesting that backbone conformational rigidity at the designed surface is important for realization of the designed function. These results can be used to improve computational design strategies, but there is still much to be learned; for example, one designed complex, which does form in experiments, was classified by all metrics as a nonbinder.