Empirical Scoring Function Research Articles

Towards an Integrated Description of Hydrogen Bonding and Dehydration: Decreasing False Positives in Virtual Screening with the HYDE Scoring Function

We developed a new empirical scoring function, HYDE, for the evaluation of protein-ligand complexes. HYDE estimates binding free energy based on two terms for dehydration and hydrogen bonding only. The essential feature of this scoring function is the integrated use of log P-derived atomic increments for the prediction of free dehydration energy and hydrogen bonding energy. Taking the dehydration of atoms within the interface into account shows that some atoms contribute favorably to the overall score, while others contribute unfavorably. For instance, hydrogen bond functions are penalized if they are dehydrated unless they can overcompensate this loss by forming a hydrogen bond with excellent geometry. The main stabilizing contribution represents the removal of apolar groups from the water: the hydrophobic effect. Initial studies using the DUD dataset show that with HYDE, there is a significant decrease in false positives, a reasonable categorization of compounds as either non-binders, weak, medium or strong binders, and in particular, there is a generally applicable and thermodynamically sensible cutoff score below which there is a high likelihood that the compound is indeed a binder.

Molecular modeling and dynamics simulations of PNP from Streptococcus agalactiae

This work describes for the first time a structural model of purine nucleoside phosphorylase from Streptococcus agalactiae ( SaPNP). PNP catalyzes the cleavage of N-ribosidic bonds of the purine ribonucleosides and 2-deoxyribonucleosides in the presence of inorganic orthophosphate as a second substrate. This enzyme is a potential target for the development of antibacterial drugs. We modeled the complexes of SaPNP with 15 different ligands in order to determine the structural basis for the specificity of these ligands against SaPNP. The application of a novel empirical scoring function to estimate the affinity of a ligand for a protein was able to identify the ligands with high affinity for PNPs. The analysis of molecular dynamics trajectory for SaPNP indicates that the functionally important motifs have a very stable structure. This new structural model together with a novel empirical scoring function opens the possibility to explorer larger library of compounds in order to identify the new inhibitors for PNPs in virtual screening projects.

A new approach for flexible protein-ligand docking based on Particle Swarm Optimisation

Particle Swarm Optimiser (PSO) uses a general-purpose, iterative, heuristic search algorithm. It considers a population of individuals to probe promising regions of the search space in an effective manner. In this context, the population of solutions is called a swarm, and the individuals are called particles. Each particle moves within the search space and retains in its memory the best position and the overall best position that has been encountered. The velocity of each particle is adjusted during each iteration toward the personal best position as well as the overall best position, thus mimicking swarm intelligence. In our recent work we have implemented PSO in a ligand-docking program. The fitness landscape of the docking program is modeled by a modified version of the scoring function X-Score [1]. X-Score is an empirical scoring function which shows a significant correlation between calculated docking scores and experimentally derived ligand geometries. Preliminary investigations show promising results in terms of speed and accuracy. Special attention during the development will be paid to a modular design of the program in order to easily implement different scoring functions as well as to perform parallel computing.

Optimizing the Signal-to-Noise Ratio of Scoring Functions for Protein−Ligand Docking

Empirical scoring functions provide estimates of the free energy of protein-ligand binding in situations when atomic-scale simulations are intractable, for example, in virtual high-throughput screening. Currently, such scoring functions are often inaccurate, and further improvements are complicated by the lack of reliable training data, the complex interplay between scoring functions and docking algorithms, and an inconsistent statistical treatment of positive and negative training data. In comparison to various other performance measures of scoring functions, "analysis of variance" provides a well-behaved objective function for optimization, which focuses on the signal-to-noise ratio of ligand-decoy discrimination. In combination with a large database of ligands and decoys, an in situ optimization of scoring function parameters was able to generate improved, target-specific scoring functions for three different proteins of pharmaceutical interest: cyclin-dependent kinase 2, the estrogen receptor, and cyclooxygenase-2. Statistical analysis of the improvements observed in "receiver-operating characteristic" curves showed that the optimized scoring functions achieved a significantly (between p < 0.0001 and p < 0.05) higher enrichment of true ligands. A scaffold dependence of the resulting binding modes was observed, which is discussed in conjunction with the rigid receptor hypothesis commonly made in protein-ligand docking. In summary, the approach described here represents a well-adapted statistical method for setting up scoring functions.

Distance Dependent Scoring Function for Describing Protein−Ligand Intermolecular Interactions

A new empirical scoring function has been developed to estimate the binding affinity of a protein-ligand complex with known three-dimensional structure. The scoring function includes a small number of physicochemical descriptors and a large number of quasi-fragmental descriptors. The first group of descriptors is chosen from the following set: (1) the number of close nonbonded contacts, (2) a score for 'metal-atom' interactions, (3) the number of flexible bonds, (4) van der Waals interaction energy, and (5) electrostatic interaction energy. A training set of 288 'protein-ligand' complexes was used to develop the scoring function. The key benefit of this approach is that it reduces the computational complexity while maintaining similar predictive ability to existing methods (average for independent test sets is 2-2.2 kcal/mol). The quasi-fragmental descriptors provide a unique and novel way of accurately representing physical interactions, compared to using physicochemical ones alone.

Customizing scoring functions for docking

Empirical scoring functions used in protein-ligand docking calculations are typically trained on a dataset of complexes with known affinities with the aim of generalizing across different docking applications. We report a novel method of scoring-function optimization that supports the use of additional information to constrain scoring function parameters, which can be used to focus a scoring function's training towards a particular application, such as screening enrichment. The approach combines multiple instance learning, positive data in the form of ligands of protein binding sites of known and unknown affinity and binding geometry, and negative (decoy) data of ligands thought not to bind particular protein binding sites or known not to bind in particular geometries. Performance of the method for the Surflex-Dock scoring function is shown in cross-validation studies and in eight blind test cases. Tuned functions optimized with a sufficient amount of data exhibited either improved or undiminished screening performance relative to the original function across all eight complexes. Analysis of the changes to the scoring function suggest that modifications can be learned that are related to protein-specific features such as active-site mobility.

Three-dimensional Structure of the EphB2 Receptor in Complex with an Antagonistic Peptide Reveals a Novel Mode of Inhibition

The Eph family of receptor tyrosine kinases has been implicated in tumorigenesis as well as pathological forms of angiogenesis. Understanding how to modulate the interaction of Eph receptors with their ephrin ligands is therefore of critical interest for the development of therapeutics to treat cancer. Previous work identified a set of 12-mer peptides that displayed moderate binding affinity but high selectivity for the EphB2 receptor. The SNEW antagonistic peptide inhibited the interaction of EphB2 with ephrinB2, with an IC50 of approximately 15 microm. To gain a better molecular understanding of how to inhibit Eph/ephrin binding, we determined the crystal structure of the EphB2 receptor in complex with the SNEW peptide to 2.3-A resolution. The peptide binds in the hydrophobic ligand-binding cleft of the EphB2 receptor, thus competing with the ephrin ligand for receptor binding. However, the binding interactions of the SNEW peptide are markedly different from those described for the TNYL-RAW peptide, which binds to the ligand-binding cleft of EphB4, indicating a novel mode of antagonism. Nevertheless, we identified a conserved structural motif present in all known receptor/ligand interfaces, which may serve as a scaffold for the development of therapeutic leads. The EphB2-SNEW complex crystallized as a homodimer, and the residues involved in the dimerization interface are similar to those implicated in mediating tetramerization of EphB2-ephrinB2 complexes. The structure of EphB2 in complex with the SNEW peptide reveals novel binding determinants that could serve as starting points in the development of compounds that modulate Eph receptor/ephrin interactions and biological activities.

Identifying bacterial species using CE–MS and SEQUEST with an empirical scoring function

CE-MS/MS analysis of proteolytic digests of bacterial cell extracts was combined with SEQUEST searching and a new scoring system to identify bacteria species in bacterial mixtures. Searches of MS/MS spectra against protein databases enabled the identification of bacterial species by the matching of the proteins associated with the corresponding species. An empirical scoring function was obtained by evaluating the SEQUEST search results of 38 samples that contained single bacterial species. The scoring by the empirical function helped move up the positive identification results from their original positions in the ranking based on Xcorr values alone. Therefore, the identification of bacteria in the samples that contained bacterial mixtures was improved. Bacterial species in 20 bacterial mixtures, including one real sample, were correctly identified by database searches and the new scoring function.

Prediction of T-Cell Epitope

The prevailing methods to predict T-cell epitopes are reviewed. Motif matching, matrix, support vector machine (SVM), and empirical scoring function methods are mainly reviewed; and the thermodynamic integration (TI) method using all-atom molecular dynamics (MD) simulation is mentioned briefly. The motif matching method appeared first and developed with the increased understanding of the characteristic structure of MHC-peptide complexes, that is, pockets aligned in the groove and corresponding residues fitting on them. This method is now becoming outdated due to the insufficiency and inaccuracy of information. The matrix method, the generalization of interaction between pockets of MHC and residues of bound peptide to all the positions in the groove, is the most prevalent one. Efficiency of calculation makes this method appropriate to scan for candidates of T-cell epitopes within whole expressed proteins in an organ or even in a body. A large amount of experimental binding data is necessary to determine a matrix. SVM is a relative of the artificial neural network, especially direct generalization of a linear Perceptron. By incorporating non-binder data and adopting encoding that reflects the physical properties of amino acids, its performance becomes quite high. Empirical scoring functions apparently seem to be founded on a physical basis. However, the estimates directly derived from the method using only structural data are far from practical use. Through regression with binding data of a series of ligands and receptors, this method predicts binding affinity with appropriate accuracy. The TI method using MD requires only structural data and a general atomic parameter, that is, force field, and hence theoretically most consistent; however, the extent of perturbation, inaccuracy of the force field, the necessity of an immense amount of calculations, and continued difficulty of sampling an adequate structure hamper the application of this method in practical use.

Comparative Evaluation of MMPBSA and XSCORE To Compute Binding Free Energy in XIAP−Peptide Complexes

Evaluation of binding free energy in receptor-ligand complexes is one of the most important challenges in theoretical drug design. Free energy is directly correlated to the thermodynamic affinity constant, and, as a first step in druglikeness, a lead compound must have this constant in the range of micro- to nanomolar activity. Many efforts have been made to calculate it by rigorous computational approaches, such as free energy perturbation or linear response approximation. However, these methods are still computationally expensive. We focus our work on XIAP, an antiapoptotic protein whose inhibition can lead to new drugs against cancer disease. We report here a comparative evaluation of two completely different methodologies to estimate binding free energy, MMPBSA (a force field based function) and XSCORE (an empirical scoring function), in seven XIAP-peptide complexes using a representative set of structures generated by previous molecular dynamics simulations. Both methods are able to predict the experimental binding free energy with acceptable errors, but if one needs to identify slight differences upon binding, MMPBSA performs better, although XSCORE is not a bad choice taking into account the low computational cost of this method.

Towards Predictive Ligand Design With Free-Energy Based Computational Methods?

The accurate prediction of ligand-biopolymer binding affinities is of general interest to medicinal chemistry, as well as to the broader field of molecular recognition. The ability to predict computationally the thermodynamics of these molecular recognition processes has been relatively weak until recently, however, continued developments on several fronts are extending the scope of applicability of these methods. The rapid growth in the number of protein-ligand structures has initially led to the development of a range of empirical scoring functions based on relatively simple descriptions of intermolecular interactions. These methods have had some success in ranking binding affinities when tuned to particular protein systems or in rather qualitative estimates of molecular fit in fast docking calculations. However, they are too unreliable for more detailed, quantitative, assessment and comparison of binding affinities. Physics-based free energy calculations are in principle more general and have the potential to be significantly more accurate. These approaches have seen steady development over many years and rely on carefully calibrated molecular energy functions (force-fields), simulations of the systems with explicit solvent, and the coming-of-age of continuum solvation models. In addition to the initially developped Free Energy Perturbation (FEP) and Thermodynamic Integration (TI) methods, new approaches include the Molecular Mechanics-Poisson-Boltzmann Surface Area (MM-PBSA) and the Linear Interaction Energy (LIE) approaches. This review concentrates on MM-PBSA and LIE, and their variants. The routine application of these calculations is becoming possible because of enhanced computational hardware and the development of a range of computational chemistry tools. This review addresses: i) the basic principles behind free energy calculations ii) recent methodological advances iii) comparisons of predicted and experimentally determined affinities iv) the uncertainties and limitations of both the computational and experimental data v) areas where progress can be made vi) the practicality of applying the methods at the different stages of the drug discovery and optimization process.

Protein-Ligand Binding Affinity Predictions by Implicit Solvent Simulations: A Tool for Lead Optimization?

Continuum electrostatics is combined with rigorous free-energy calculations in an effort to deliver a reliable and efficient method for in silico lead optimization. The methodology is tested by calculation of the relative binding free energies of a set of inhibitors of neuraminidase, cyclooxygenase2, and cyclin-dependent kinase 2. The calculated free energies are compared to the results obtained with explicit solvent simulations and empirical scoring functions. For cyclooxygenase2, deficiencies in the continuum electrostatics theory are identified and corrected with a modified simulation protocol. For neuraminidase, it is shown that a continuum representation of the solvent leads to markedly different protein-ligand interactions compared to the explicit solvent simulations, and a reconciliation of the two protocols is problematic. Cyclin-dependent kinase 2 proves more challenging, and none of the methods employed in this study yield high quality predictions. Despite the differences observed, for these systems, the use of an implicit solvent framework to predict the ranking of congeneric inhibitors to a protein is shown to be faster, as accurate or more accurate than the explicit solvent protocol, and superior to empirical scoring schemes.

Inhibitor Binding to the Plasmepsin IV Aspartic Protease from Plasmodium falciparum

Plasmepsin IV (Plm IV) is one of the aspartic proteases present in the food vacuole of the malaria parasite Plasmodium falciparum involved in host hemoglobin degradation by the parasite. Using a series of previously synthesized plasmepsin inhibitors [Ersmark, K., et al. (2005) J. Med. Chem. 48, 6090-106], we report here experimental data and theoretical analysis of their inhibitory activity toward Plm IV. All compounds share a 1,2-dihydroxyethylene unit as the transition state mimic. They possess symmetric P1 and P1' side chains and either a diacylhydrazine, a five-membered oxadiazole ring, or a retroamide at the P2 and P2' positions. Experimental binding affinities are compared to those predicted by the linear interaction energy (LIE) method and an empirical scoring function, using both a crystal structure and a homology model for the enzyme. Molecular dynamics (MD) simulations of the modeled complexes allow a rational interpretation of the structural determinants for inhibitor binding. A ligand bearing a P2 and P2' symmetric oxadiazole which is devoid of amide bonds is identified both experimentally and theoretically as the most potent inhibitor of Plm IV. For the P2 and P2' asymmetric compounds, the results are consistent with earlier predictions regarding the mode of binding of this class of inhibitors to Plm II. Theoretical estimation of selectivity for some compounds is also reported. Significant features of the Plm IV binding pocket are discussed in comparison to related enzymes, and the results obtained here should be helpful for further optimization of inhibitors.

Are Automated Molecular Dynamics Simulations and Binding Free Energy Calculations Realistic Tools in Lead Optimization? An Evaluation of the Linear Interaction Energy (LIE) Method

An extensive evaluation of the linear interaction energy (LIE) method for the prediction of binding affinity of docked compounds has been performed, with an emphasis on its applicability in lead optimization. An automated setup is presented, which allows for the use of the method in an industrial setting. Calculations are performed for four realistic examples, retinoic acid receptor gamma, matrix metalloprotease 3, estrogen receptor alpha, and dihydrofolate reductase, focusing on different aspects of the procedure. The obtained LIE models are evaluated in terms of the root-mean-square (RMS) errors from experimental binding free energies and the ability to rank compounds appropriately. The results are compared to the best empirical scoring function, selected from a set of 10 scoring functions. In all cases, good LIE models can be obtained in terms of free-energy RMS errors, although reasonable ranking of the ligands of dihydrofolate reductase proves difficult for both the LIE method and scoring functions. For the other proteins, the LIE model results in better predictions than the best performing scoring function. These results indicate that the LIE approach, as a tool to evaluate docking results, can be a valuable asset in computational lead optimization programs.

Protein Structure Evaluation using an All-Atom Energy Based Empirical Scoring Function

Arriving at the native conformation of a polypeptide chain characterized by minimum most free energy is a problem of long standing interest in protein structure prediction endeavors. Owing to the computational requirements in developing free energy estimates, scoring functions—energy based or statistical—have received considerable renewed attention in recent years for distinguishing native structures of proteins from non-native like structures. Several cleverly designed decoy sets, CASP (Critical Assessment of Techniques for Protein Structure Prediction) structures and homology based internet accessible three dimensional model builders are now available for validating the scoring functions. We describe here an all-atom energy based empirical scoring function and examine its performance on a wide series of publicly available decoys. Barring two protein sequences where native structure is ranked second and seventh, native is identified as the lowest energy structure in 67 protein sequences from among 61,659 decoys belonging to 12 different decoy sets. We further illustrate a potential application of the scoring function in bracketing native-like structures of two small mixed alpha/beta globular proteins starting from sequence and secondary structural information. The scoring function has been web enabled at www.scfbio-iitd.res.in/utility/proteomics/energy.jsp

Prediction of transmembrane helix orientation in polytopic membrane proteins

BackgroundMembrane proteins compose up to 30% of coding sequences within genomes. However, their structure determination is lagging behind compared with soluble proteins due to the experimental difficulties. Therefore, it is important to develop reliable computational methods to predict structures of membrane proteins.ResultsWe present a method for prediction of the TM helix orientation, which is an essential step in ab initio modeling of membrane proteins. Our method is based on a canonical model of the heptad repeat originally developed for coiled coils. We identify the helical surface patches that interface with lipid molecules at an accuracy of about 88% from the sequence information alone, using an empirical scoring function LIPS (LIPid-facing Surface), which combines lipophilicity and conservation of residues in the helix. We test and discuss results of prediction of helix-lipid interfaces on 162 transmembrane helices from 18 polytopic membrane proteins and present predicted orientations of TM helices in TRPV1 channel. We also apply our method to two structures of homologous cytochrome b6f complexes and find discrepancy in the assignment of TM helices from subunits PetG, PetN and PetL. The results of LIPS calculations and analysis of packing and H-bonding interactions support the helix assignment found in the cytochrome b6f structure from green alga but not the assignment of TM helices in the cyanobacterium b6f structure.ConclusionLIPS calculations can be used for the prediction of helix orientation in ab initio modeling of polytopic membrane proteins. We also show with the example of two cytochrome b6f structures that our method can identify questionable helix assignments in membrane proteins. The LIPS server is available online at .

An all atom energy based computational protocol for predicting binding affinities of protein–ligand complexes

We report here a computationally fast protocol for predicting binding affinities of non-metallo protein–ligand complexes. The protocol builds in an all atom energy based empirical scoring function comprising electrostatics, van der Waals, hydrophobicity and loss of conformational entropy of protein side chains upon ligand binding. The method is designed to ensure transferability across diverse systems and has been validated on a heterogenous dataset of 161 complexes consisting of 55 unique protein targets. The scoring function trained on a dataset of 61 complexes yielded a correlation of r = 0.92 for the predicted binding free energies against the experimental binding affinities. Model validation and parameter analysis studies ensure the predictive ability of the scoring function. When tested on the remaining 100 protein–ligand complexes a correlation of r = 0.92 was recovered. The high correlation obtained underscores the potential applicability of the methodology in drug design endeavors. The scoring function has been web enabled at www.scfbio-iitd.res.in/software/drugdesign/bappl.jsp as binding affinity prediction of protein–ligand (BAPPL) server.

Representative brain selection using a group-specific tissue probability map

To determine the anatomy of a standard brain reflecting well-defined group characteristics based on probabilistic information from group-specific anatomical variations. We constructed a group-specific tissue probabilistic map for 20 subjects and used it to extract voxel-wise probabilistic information for each subject through regional spatial normalization using Automated Image Registration software (AIR 5.2.5). Extracted probabilistic information was then used to determine standard properties of the subjects. For comparison, we employed an empirical scoring function - a measure of entropy - in ordering the data set. A brain with minimum entropy was then selected for a group standard. The evaluation of our proposed method was performed using two different selection schemes: deformation analysis and similarity index measurements. This method showed highly correlated result with previous method by Kochunov et al., with fewer computational tasks. This method can thus be used to determine an appropriate standard model to compare with disease-affected brains.

LigScore: a novel scoring function for predicting binding affinities

We present two new empirical scoring functions, LigScore1 and LigScore2, that attempt to accurately predict the binding affinity between ligand molecules and their protein receptors. The LigScore functions consist of three distinct terms that describe the van der Waals interaction, the polar attraction between the ligand and protein, and the desolvation penalty attributed to the binding of the polar ligand atoms to the protein and vice versa. Utilizing a regression approach on a data set of 118 protein–ligand complexes we have obtained a linear equation, LigScore2, using these three descriptors. LigScore2 has good predictability with regard to experimental p K i values yielding a correlation coefficient, r 2, of 0.75 and a standard deviation of 1.04 over the training data set, which consists of a diverse set of proteins that span more than seven protein families.

Free Energy of Ligand Binding to Protein: Evaluation of the Contribution of Water Molecules by Computational Methods

One of the more challenging issues in medicinal chemistry is the computation of the free energy of ligand binding to macromolecular targets. This allows for the screening of libraries of chemicals for fast and inexpensive identification of lead compounds. Many attempts have been made and several algorithms have been developed for this purpose. Whereas enthalpic contributions are evaluated using methods and equations for which there is a reasonable consensus among researchers, the entropic contribution is evaluated using very different, and, in some cases, very approximate methods, or it is entirely ignored. Entropic contributions are of primary importance in the formation of many ligand-protein complexes, as well as in protein folding. The hydrophobic interaction, associated with the release of water molecules from the protein active site and the ligand, plays a significant role in complex formation, predominantly contributing to the total entropy change and, in some cases, to the total free energy of binding. There are distinct approaches for the evaluation of the contribution of water molecules to the free energy of binding based on Newtonian mechanics force fields, multi-parameter empirical scoring functions and experimental force fields. This review describes these methods -- discussing both their advantages and limitations. Particular emphasis will be placed on HINT (Hydropatic INTeractions), a "natural" force field that takes into account in a unified way enthalpic and entropic contributions of all interacting atoms in protein-ligand complexes, including released and structured water molecules. As a case-study, the contribution of water molecules to the binding free energy of HIV-1 protease inhibitors is evaluated.