Docking prediction using biological information, ZDOCK sampling technique, and clustering guided by the DFIRE statistical energy function

Rational Prediction of PROTAC-Compatible Protein-Protein Interfaces by Molecular Docking

Proteolysis targeting chimeras (PROTACs) are heterobifunctional ligands that mediate the interaction between a protein target and an E3 ligase, resulting in a ternary complex, whose interaction with the ubiquitination machinery leads to target degradation. This technology is emerging as an exciting new avenue for therapeutic development, with several PROTACs currently undergoing clinical trials targeting cancer. Here, we describe a general and computationally efficient methodology combining restraint-based docking, energy-based rescoring, and a filter based on the minimal solvent-accessible surface distance to produce PROTAC-compatible PPIs suitable for when there is no a priori known PROTAC ligand. In a benchmark employing a manually curated data set of 13 ternary complex crystals, we achieved an accuracy of 92% when starting from bound structures and 77% when starting from unbound structures, respectively. Our method only requires that the ligand-bound structures of the monomeric forms of the E3 ligase and target proteins be given to run, making it general, accurate, and highly efficient, with the ability to impact early-stage PROTAC-based drug design campaigns where no structural information about the ternary complex structure is available.

TRScore: a three-dimensional RepVGG-based scoring method for ranking protein docking models

MOTIVATION

Protein-protein interactions (PPI) play important roles in cellular activities. Due to the technical difficulty and high cost of experimental methods, there are considerable interests towards the development of computational approaches, such as protein docking, to decipher PPI patterns. One of the important and difficult aspects in protein docking is recognizing near-native conformations from a set of decoys, but unfortunately traditional scoring functions still suffer from limited accuracy. Therefore, new scoring methods are pressingly needed in methodological and/or practical implications.

RESULTS

We present a new deep learning-based scoring method for ranking protein-protein docking models based on a three-dimensional (3D) RepVGG network, named TRScore. To recognize near-native conformations from a set of decoys, TRScore voxelizes the protein-protein interface into a 3D grid labeled by the number of atoms in different physicochemical classes. Benefiting from the deep convolutional RepVGG architecture, TRScore can effectively capture the subtle differences between energetically favorable near-native models and unfavorable non-native decoys without needing extra information. TRScore was extensively evaluated on diverse test sets including protein-protein docking benchmark 5.0 update set, DockGround decoy set, as well as realistic CAPRI decoy set, and overall obtained a significant improvement over existing methods in cross validation and independent evaluations.

AVAILABILITY

Codes available at: https://github.com/BioinformaticsCSU/TRScore.

Molecular Docking at a Glance

The current chapter introduces different aspects of molecular docking technique in order to give an overview to the readers about the topics which will be dealt with throughout this volume. Like many other fields of science, molecular docking studies has experienced a lagging period of slow and steady increase in terms of acquiring attention of scientific community as well as its frequency of application, followed by a pronounced era of exponential expansion in theory, methodology, areas of application and performance due to developments in related technologies such as computational resources and theoretical as well as experimental biophysical methods. In the following sections the evolution of molecular docking will be reviewed and its different components including methods, search algorithms, scoring functions, validation of the methods, and area of applications plus few case studies will be touched briefly.

A simple, high-throughput modeling approach reveals insights into the mechanism of gametophytic self-incompatibility

Specificity in the GSI response results from the S-haplotype-specific molecular interaction of S-locus F-box (SLF/SFB) and SRNase proteins in the self-incompatibility locus (S-locus). The answer to the question of how these two components of the S-locus (SRNase and SLF/SFB) interact has been gathered from several models. Since there is not enough evidence as to which one is the definitive model, none of them can be ruled out. Despite the identification of interacting protein elements, the mechanism by which SLF/SFB and SRNase interact to differently trigger the self-incompatibility among families and subfamilies remain uncertain. The high-throughput modeling approach demonstrates structural visions into the possible existence of a Collaborative Non-Self Recognition model in apple. These findings postulate several prospects for future investigation providing useful information to guide the implementation of breeding strategies.

Network mapping among the functional domains of Chikungunya virus nonstructural proteins

Formation of virus specific replicase complex is among the most important steps that determines the fate of viral transcription and replication during Chikungunya virus (CHIKV) infection. In the present study, the authors have computationally generated a 3D structure of CHIKV late replicase complex on the basis of the interactions identified among the domains of CHIKV nonstructural proteins (nsPs) which make up the late replicase complex. The interactions among the domains of CHIKV nsPs were identified using systems such as pull down, protein interaction ELISA, and yeast two-hybrid. The structures of nsPs were generated using I-TASSER and the biological assembly of the replicase complex was determined using ZRANK and RDOCK. A total of 36 interactions among the domains and full length proteins were tested and 12 novel interactions have been identified. These interactions included the homodimerization of nsP1 and nsP4 through their respective C-ter domains; the associations of nsP2 helicase domain and C-ter domain of nsP4 with methyltransferase and membrane binding domains of nsP1; the interaction of nsP2 protease domain with C-ter domain of nsP4; and the interaction of nsP3 macro and alphavirus unique domains with the C-ter domain of nsP1. The novel interactions identified in the current study form a network of organized associations that suggest the spatial arrangement of nsPs in the late replicase complex of CHIKV.

How to use not-always-reliable binding site information in protein-protein docking prediction

In many protein-protein docking algorithms, binding site information is used to help predicting the protein complex structures. Using correct and accurate binding site information can increase protein-protein docking success rate significantly. On the other hand, using wrong binding sites information should lead to a failed prediction, or, at least decrease the success rate. Recently, various successful theoretical methods have been proposed to predict the binding sites of proteins. However, the predicted binding site information is not always reliable, sometimes wrong binding site information could be given. Hence there is a high risk to use the predicted binding site information in current docking algorithms. In this paper, a softly restricting method (SRM) is developed to solve this problem. By utilizing predicted binding site information in a proper way, the SRM algorithm is sensitive to the correct binding site information but insensitive to wrong information, which decreases the risk of using predicted binding site information. This SRM is tested on benchmark 3.0 using purely predicted binding site information. The result shows that when the predicted information is correct, SRM increases the success rate significantly; however, even if the predicted information is completely wrong, SRM only decreases success rate slightly, which indicates that the SRM is suitable for utilizing predicted binding site information.

Sampling and scoring: a marriage made in heaven

Most structure prediction algorithms consist of initial sampling of the conformational space, followed by rescoring and possibly refinement of a number of selected structures. Here we focus on protein docking, and show that while decoupling sampling and scoring facilitates method development, integration of the two steps can lead to substantial improvements in docking results. Since decoupling is usually achieved by generating a decoy set containing both non-native and near-native docked structures, which can be then used for scoring function construction, we first review the roles and potential pitfalls of decoys in protein-protein docking, and show that some type of decoys are better than others for method development. We then describe three case studies showing that complete decoupling of scoring from sampling is not the best choice for solving realistic docking problems. Although some of the examples are based on our own experience, the results of the CAPRI docking and scoring experiments also show that performing both sampling and scoring generally yields better results than scoring the structures generated by all predictors. Next we investigate how the selection of training and decoy sets affects the performance of the scoring functions obtained. Finally, we discuss pathways to better alignment of the two steps, and show some algorithms that achieve a certain level of integration. Although we focus on protein-protein docking, our observations most likely also apply to other conformational search problems, including protein structure prediction and the docking of small molecules to proteins.

Optimized grid-based protein-protein docking as a global search tool followed by incorporating experimentally derivable restraints

Unbound protein docking, or the computational prediction of the structure of a protein complex from the structures of its separated components, is of importance but still challenging. A practical approach toward reliable results for unbound docking is to incorporate experimentally derived information with computation. To this end, truly systematic search of the global docking space is desirable. The fast Fourier transform (FFT) docking is a systematic search method with high computational efficiency. However, by using FFT to perform unbound docking, possible conformational changes upon binding must be treated implicitly. To better accommodate the implicit treatment of conformational flexibility, we develop a rational approach to optimize "softened" parameters for FFT docking. In connection with the increased "softness" of the parameters in this global search step, we use a revised rule to select candidate models from the search results. For complexes designated as of low and medium difficulty for unbound docking, these adaptations of the original FTDOCK program lead to substantial improvements of the global search results. Finally, we show that models resulted from FFT-based global search can be further filtered with restraints derivable from nuclear magnetic resonance (NMR) chemical shift perturbation or mutagenesis experiments, leading to a small set of models that can be feasibly refined and evaluated using computationally more expensive methods and that still include high-ranking near-native conformations.

DEVELOPMENT OF AN AFFINITY EVALUATION AND PREDICTION SYSTEM BY USING THE SHAPE COMPLEMENTARITY CHARACTERISTIC BETWEEN PROTEINS

A system was developed to evaluate and predict the interaction between protein pairs by using the widely used shape complementarity search method as the algorithm for docking simulations between the proteins. This system, which we call the affinity evaluation and prediction (AEP) system, was used to evaluate the interaction between 20 protein pairs. The system first executes a "round robin" shape complementarity search of the target protein group, and evaluates the interaction of the complex structures obtained by shape complementarity search. These complex structures are selected by using a statistical procedure that we developed called "grouping". At a low prevalence of 5.0%, our AEP system predicted protein–protein interaction with 65.0% recall, 15.1% precision, 80.0% accuracy, and had an area under the curve (AUC) of 0.74. By optimizing the grouping process, our AEP system successfully predicted 13 protein pairs (among 20 pairs) that were biologically significant combinations. Our ultimate goal is to construct an affinity database that will provide crucial information obtained using our AEP system to cell biologists and drug designers.

Refining near-native protein-protein docking decoys by local resampling and energy minimization

How to refine a near-native structure to make it closer to its native conformation is an unsolved problem in protein-structure and protein-protein complex-structure prediction. In this article, we first test several scoring functions for selecting locally resampled near-native protein-protein docking conformations and then propose a computationally efficient protocol for structure refinement via local resampling and energy minimization. The proposed method employs a statistical energy function based on a Distance-scaled Ideal-gas REference state (DFIRE) as an initial filter and an empirical energy function EMPIRE (EMpirical Protein-InteRaction Energy) for optimization and re-ranking. Significant improvement of final top-1 ranked structures over initial near-native structures is observed in the ZDOCK 2.3 decoy set for Benchmark 1.0 (74% whose global rmsd reduced by 0.5 A or more and only 7% increased by 0.5 A or more). Less significant improvement is observed for Benchmark 2.0 (38% versus 33%). Possible reasons are discussed.

A simple reference state makes a significant improvement in near-native selections from structurally refined docking decoys

Near-native selections from docking decoys have proved challenging especially when unbound proteins are used in the molecular docking. One reason is that significant atomic clashes in docking decoys lead to poor predictions of binding affinities of near native decoys. Atomic clashes can be removed by structural refinement through energy minimization. Such an energy minimization, however, will lead to an unrealistic bias toward docked structures with large interfaces. Here, we extend an empirical energy function developed for protein design to protein-protein docking selection by introducing a simple reference state that removes the unrealistic dependence of binding affinity of docking decoys on the buried solvent accessible surface area of interface. The energy function called EMPIRE (EMpirical Protein-InteRaction Energy), when coupled with a refinement strategy, is found to provide a significantly improved success rate in near native selections when applied to RosettaDock and refined ZDOCK docking decoys. Our work underlines the importance of removing nonspecific interactions from specific ones in near native selections from docking decoys.

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.

The development of an affinity evaluation and prediction system by using protein-protein docking simulations and parameter tuning

A system was developed to evaluate and predict the interaction between protein pairs by using the widely used shape complementarity search method as the algorithm for docking simulations between the proteins. We used this system, which we call the affinity evaluation and prediction (AEP) system, to evaluate the interaction between 20 protein pairs. The system first executes a “round robin” shape complementarity search of the target protein group, and evaluates the interaction between the complex structures obtained by the search. These complex structures are selected by using a statistical procedure that we developed called ‘grouping’. At a prevalence of 5.0%, our AEP system predicted protein–protein interactions with a 50.0% recall, 55.6% precision, 95.5% accuracy, and an F-measure of 0.526. By optimizing the grouping process, our AEP system successfully predicted 10 protein pairs (among 20 pairs) that were biologically relevant combinations. Our ultimate goal is to construct an affinity database that will provide cell biologists and drug designers with crucial information obtained using our AEP system.

Homology modelling of protein-protein complexes: a simple method and its possibilities and limitations

Background

Structure-based computational methods are needed to help identify and characterize protein-protein complexes and their function. For individual proteins, the most successful technique is homology modelling. We investigate a simple extension of this technique to protein-protein complexes. We consider a large set of complexes of known structures, involving pairs of single-domain proteins. The complexes are compared with each other to establish their sequence and structural similarities and the relation between the two. Compared to earlier studies, a simpler dataset, a simpler structural alignment procedure, and an additional energy criterion are used. Next, we compare the Xray structures to models obtained by threading the native sequence onto other, homologous complexes. An elementary requirement for a successful energy function is to rank the native structure above any threaded structure. We use the DFIRE_β energy function, whose quality and complexity are typical of the models used today. Finally, we compare near-native models to distinctly non-native models.

Results

If weakly stable complexes are excluded (defined by a binding energy cutoff), as well as a few unusual complexes, a simple homology principle holds: complexes that share more than 35% sequence identity share similar structures and interaction modes; this principle was less clearcut in earlier studies. The energy function was then tested for its ability to identify experimental structures among sets of decoys, produced by a simple threading procedure. On average, the experimental structure is ranked above 92% of the alternate structures. Thus, discrimination of the native structure is good but not perfect. The discrimination of near-native structures is fair. Typically, a single, alternate, non-native binding mode exists that has a native-like energy. Some of the associated failures may correspond to genuine, alternate binding modes and/or native complexes that are artefacts of the crystal environment. In other cases, additional model filtering with more sophisticated tools is needed.

Conclusion

The results suggest that the simple modelling procedure applied here could help identify and characterize protein-protein complexes. The next step is to apply it on a genomic scale.

An iterative knowledge-based scoring function for protein-protein recognition

Using an efficient iterative method, we have developed a distance-dependent knowledge-based scoring function to predict protein-protein interactions. The function, referred to as ITScore-PP, was derived using the crystal structures of a training set of 851 protein-protein dimeric complexes containing true biological interfaces. The key idea of the iterative method for deriving ITScore-PP is to improve the interatomic pair potentials by iteration, until the pair potentials can distinguish true binding modes from decoy modes for the protein-protein complexes in the training set. The iterative method circumvents the challenging reference state problem in deriving knowledge-based potentials. The derived scoring function was used to evaluate the ligand orientations generated by ZDOCK 2.1 and the native ligand structures on a diverse set of 91 protein-protein complexes. For the bound test cases, ITScore-PP yielded a success rate of 98.9% if the top 10 ranked orientations were considered. For the more realistic unbound test cases, the corresponding success rate was 40.7%. Furthermore, for faster orientational sampling purpose, several residue-level knowledge-based scoring functions were also derived following the similar iterative procedure. Among them, the scoring function that uses the side-chain center of mass (SCM) to represent a residue, referred to as ITScore-PP(SCM), showed the best performance and yielded success rates of 71.4% and 30.8% for the bound and unbound cases, respectively, when the top 10 orientations were considered. ITScore-PP was further tested using two other published protein-protein docking decoy sets, the ZDOCK decoy set and the RosettaDock decoy set. In addition to binding mode prediction, the binding scores predicted by ITScore-PP also correlated well with the experimentally determined binding affinities, yielding a correlation coefficient of R = 0.71 on a test set of 74 protein-protein complexes with known affinities. ITScore-PP is computationally efficient. The average run time for ITScore-PP was about 0.03 second per orientation (including optimization) on a personal computer with 3.2 GHz Pentium IV CPU and 3.0 GB RAM. The computational speed of ITScore-PP(SCM) is about an order of magnitude faster than that of ITScore-PP. ITScore-PP and/or ITScore-PP(SCM) can be combined with efficient protein docking software to study protein-protein recognition.

Incorporating biochemical information and backbone flexibility in RosettaDock for CAPRI rounds 6-12

In CAPRI rounds 6-12, RosettaDock successfully predicted 2 of 5 unbound-unbound targets to medium accuracy. Improvement over the previous method was achieved with computational mutagenesis to select decoys that match the energetics of experimentally determined hot spots. In the case of Target 21, Orc1/Sir1, this resulted in a successful docking prediction where RosettaDock alone or with simple site constraints failed. Experimental information also helped limit the interacting region of TolB/Pal, producing a successful prediction of Target 26. In addition, we docked multiple loop conformations for Target 20, and we developed a novel flexible docking algorithm to simultaneously optimize backbone conformation and rigid-body orientation to generate a wide diversity of conformations for Target 24. Continued challenges included docking of homology targets that differ substantially from their template (sequence identity <50%) and accounting for large conformational changes upon binding. Despite a larger number of unbound-unbound and homology model binding targets, Rounds 6-12 reinforced that RosettaDock is a powerful algorithm for predicting bound complex structures, especially when combined with experimental data.

Recognizing protein-protein interfaces with empirical potentials and reduced amino acid alphabets

Background

In structural genomics, an important goal is the detection and classification of protein–protein interactions, given the structures of the interacting partners. We have developed empirical energy functions to identify native structures of protein–protein complexes among sets of decoy structures. To understand the role of amino acid diversity, we parameterized a series of functions, using a hierarchy of amino acid alphabets of increasing complexity, with 2, 3, 4, 6, and 20 amino acid groups. Compared to previous work, we used the simplest possible functional form, with residue–residue interactions and a stepwise distance-dependence. We used increased computational ressources, however, constructing 290,000 decoys for 219 protein–protein complexes, with a realistic docking protocol where the protein partners are flexible and interact through a molecular mechanics energy function. The energy parameters were optimized to correctly assign as many native complexes as possible. To resolve the multiple minimum problem in parameter space, over 64000 starting parameter guesses were tried for each energy function. The optimized functions were tested by cross validation on subsets of our native and decoy structures, by blind tests on series of native and decoy structures available on the Web, and on models for 13 complexes submitted to the CAPRI structure prediction experiment.

Results

Performance is similar to several other statistical potentials of the same complexity. For example, the CAPRI target structure is correctly ranked ahead of 90% of its decoys in 6 cases out of 13. The hierarchy of amino acid alphabets leads to a coherent hierarchy of energy functions, with qualitatively similar parameters for similar amino acid types at all levels. Most remarkably, the performance with six amino acid classes is equivalent to that of the most detailed, 20-class energy function.

Conclusion

This suggests that six carefully chosen amino acid classes are sufficient to encode specificity in protein–protein interactions, and provide a starting point to develop more complicated energy functions.

FastContact: a free energy scoring tool for protein-protein complex structures

‘FastContact’ is a server that estimates the direct electrostatic and desolvation interaction free energy between two proteins in units of kcal/mol. Users submit two proteins in PDB format, and the output is emailed back to the user in three files: one output file, and the two processed proteins. Besides the electrostatic and desolvation free energy, the server reports residue contact free energies that rapidly highlight the hotspots of the interaction and evaluates the van der Waals interaction using CHARMm. Response time is ∼1 min. The server has been successfully tested and validated, scoring refined complex structures and blind sets of docking decoys, as well as proven useful predicting protein interactions. ‘FastContact’ offers unique capabilities from biophysical insights to scoring and identifying important contacts.

A free-rotating and self-avoiding chain model for deriving statistical potentials based on protein structures

Statistical potentials have been widely used in protein studies despite the much-debated theoretical basis. In this work, we have applied two physical reference states for deriving the statistical potentials based on protein structure features to achieve zero interaction and orthogonalization. The free-rotating chain-based potential applies a local free-rotating chain reference state, which could theoretically be described by the Gaussian distribution. The self-avoiding chain-based potential applies a reference state derived from a database of artificial self-avoiding backbones generated by Monte Carlo simulation. These physical reference states are independent of known protein structures and are based solely on the analytical formulation or simulation method. The new potentials performed better and yielded higher Z-scores and success rates compared to other statistical potentials. The end-to-end distance distribution produced by the self-avoiding chain model was similar to the distance distribution of protein atoms in structure database. This fact may partly explain the basis of the reference states that depend on the atom pair frequency observed in the protein database. The current study showed that a more physical reference model improved the performance of statistical potentials in protein fold recognition, which could also be extended to other types of applications.

A protein-specifically adapted scoring function for the reranking of docking solutions

In this work, we developed a protein-specifically adapted scoring function and applied it to the reranking of protein-protein docking solutions generated with a conventional docking program. The approach was validated using experimentally determined structures of the bacterial HPr-protein in complex with four structurally nonhomologous binding partners as an example. A sufficiently large data basis for the generation of protein-specifically adapted pair potentials was generated by modeling all orthologous complexes for each type of interaction resulting in a total of 224 complexes. The parameters for potential generation were systematically varied and resulted in a total of 66,132 different scoring functions that were tested for their ability of successful reranking of 1000 docking solutions generated from modeled structures of the unbound binding partners. Parameters that proved critical for the generation of good scoring functions were the distance cutoff used for the generation of the pair potential, and an additional cutoff that allows a proper weighting of conserved and nonconserved contacts in the interface. Compared to the original scoring function, application of this novel type of scoring functions resulted in a significant accumulation of acceptable docking solutions within the first 10 ranks. Depending on the type of complex investigated one to five acceptable complex geometries are found among the 10 highest-ranked solutions and for three of the four systems tested, an acceptable solution was placed on the first rank.

A new protein-protein docking scoring function based on interface residue properties

MOTIVATION

Protein-protein complexes are known to play key roles in many cellular processes. However, they are often not accessible to experimental study because of their low stability and difficulty to produce the proteins and assemble them in native conformation. Thus, docking algorithms have been developed to provide an in silico approach of the problem. A protein-protein docking procedure traditionally consists of two successive tasks: a search algorithm generates a large number of candidate solutions, and then a scoring function is used to rank them.

RESULTS

To address the second step, we developed a scoring function based on a Voronoï tessellation of the protein three-dimensional structure. We showed that the Voronoï representation may be used to describe in a simplified but useful manner, the geometric and physico-chemical complementarities of two molecular surfaces. We measured a set of parameters on native protein-protein complexes and on decoys, and used them as attributes in several statistical learning procedures: a logistic function, Support Vector Machines (SVM), and a genetic algorithm. For the later, we used ROGER, a genetic algorithm designed to optimize the area under the receiver operating characteristics curve. To further test the scores derived with ROGER, we ranked models generated by two different docking algorithms on targets of a blind prediction experiment, improving in almost all cases the rank of native-like solutions.

AVAILABILITY

http://genomics.eu.org/spip/-Bioinformatics-tools-

Protein binding site prediction using an empirical scoring function

Most biological processes are mediated by interactions between proteins and their interacting partners including proteins, nucleic acids and small molecules. This work establishes a method called PINUP for binding site prediction of monomeric proteins. With only two weight parameters to optimize, PINUP produces not only 42.2% coverage of actual interfaces (percentage of correctly predicted interface residues in actual interface residues) but also 44.5% accuracy in predicted interfaces (percentage of correctly predicted interface residues in the predicted interface residues) in a cross validation using a 57-protein dataset. By comparison, the expected accuracy via random prediction (percentage of actual interface residues in surface residues) is only 15%. The binding sites of the 57-protein set are found to be easier to predict than that of an independent test set of 68 proteins. The average coverage and accuracy for this independent test set are 30.5 and 29.4%, respectively. The significant gain of PINUP over expected random prediction is attributed to (i) effective residue-energy score and accessible-surface-area-dependent interface-propensity, (ii) isolation of functional constraints contained in the conservation score from the structural constraints through the combination of residue-energy score (for structural constraints) and conservation score and (iii) a consensus region built on top-ranked initial patches.

Scoring a diverse set of high-quality docked conformations: A metascore based on electrostatic and desolvation interactions

Predicting protein-protein interactions involves sampling and scoring docked conformations. Barring some large structural rearrangement, rapidly sampling the space of docked conformations is now a real possibility, and the limiting step for the successful prediction of protein interactions is the scoring function used to reduce the space of conformations from billions to a few, and eventually one high affinity complex. An atomic level free-energy scoring function that estimates in units of kcal/mol both electrostatic and desolvation interactions (plus van der Waals if appropriate) of protein-protein docked conformations is used to rerank the blind predictions (860 in total) submitted for six targets to the community-wide Critical Assessment of PRediction of Interactions (CAPRI; http://capri.ebi.ac.uk). We found that native-like models often have varying intermolecular contacts and atom clashes, making unlikely that one can construct a universal function that would rank all these models as native-like. Nevertheless, our scoring function is able to consistently identify the native-like complexes as those with the lowest free energy for the individual models of 16 (out of 17) human predictors for five of the targets, while at the same time the modelers failed to do so in more than half of the cases. The scoring of high-quality models developed by a wide variety of methods and force fields confirms that electrostatic and desolvation forces are the dominant interactions determining the bound structure. The CAPRI experiment has shown that modelers can predict valuable models of protein-protein complexes, and improvements in scoring functions should soon solve the docking problem for complexes whose backbones do not change much upon binding. A scoring server and programs are available at http://structure.pitt.edu.