Modeling protein-ligand complexes

Identification of Phenolic Compounds from Nettle as New Candidate Inhibitors of Main Enzymes Responsible on Type-II Diabetes

BACKGROUND

In medicinal chemistry, the discovery of small organic molecules that can be optimized and lead to a future drug capable of effectively modulating the biological activity of a therapeutic target remains a major challenge. Because of the harmful secondary effects of synthesized therapeutic molecules, the development of research has been oriented towards phytomedicines. Phenolic compounds from medicinal plants are constantly explored for new therapeutic use.

METHODS

In this paper, we studied interactions between main enzymes responsible for causing type 2 diabetes mellitus (T2DM) and phenolic compounds from nettle (Urtica dioica L.) using molecular Docking with Molecular Operating Environment Software (MOE).

RESULTS

Docking results show a common molecule (secoisolariciresinol), which may form stable complexes with depeptidyl peptidase 4 (DPP-4), alpha-amylase and beta-glucosidase with binding energy of -7.04732084 kcal/mol, -3.82946181 kcal/mol and -4.16077089 kcal/mol respectively. Besides secoisolariciresinol, other phenolic compounds give better docking score than the original co-crystallized ligand for alpha-amylase (PDB ID 5U3A) and beta-glucosidase (PDB ID 1OGS).

CONCLUSION

The obtained results are promising for the discovery of new alpha-amylase and betaglucosidase inhibitors. This study also confirms the folk use of nettle as antidiabetic agent.

Inhibition of protein interactions: co-crystalized protein-protein interfaces are nearly as good as holo proteins in rigid-body ligand docking

Modulating protein interaction pathways may lead to the cure of many diseases. Known protein-protein inhibitors bind to large pockets on the protein-protein interface. Such large pockets are detected also in the protein-protein complexes without known inhibitors, making such complexes potentially druggable. The inhibitor-binding site is primary defined by the side chains that form the largest pocket in the protein-bound conformation. Low-resolution ligand docking shows that the success rate for the protein-bound conformation is close to the one for the ligand-bound conformation, and significantly higher than for the apo conformation. The conformational change on the protein interface upon binding to the other protein results in a pocket employed by the ligand when it binds to that interface. This proof-of-concept study suggests that rather than using computational pocket-opening procedures, one can opt for an experimentally determined structure of the target co-crystallized protein-protein complex as a starting point for drug design.

Zika virus drug targets: a missing link in drug design and discovery – a route map to fill the gap

This review depicts anin silicoroute map for ZIKV drug discovery, thus revealing novel potential inhibitors of viral replication.

Prediction of SAMPL2 aqueous solvation free energies and tautomeric ratios using the SM8, SM8AD, and SMD solvation models

We applied the solvation models SM8, SM8AD, and SMD in combination with the Minnesota M06-2X density functional to predict vacuum-water transfer free energies (Task 1) and tautomeric ratios in aqueous solution (Task 2) for the SAMPL2 test set. The bulk-electrostatic contribution to the free energy of solvation is treated as follows: SM8 employs the generalized Born model with the Coulomb field approximation, SM8AD employs the generalized Born approximation with asymmetric descreening, and SMD solves the nonhomogeneous Poisson equation. The non-bulk-electrostatic contribution arising from short-range interactions between the solute and solvent molecules in the first solvation shell is treated as a sum of terms that are products of geometry-dependent atomic surface tensions and solvent-accessible surface areas of the individual atoms of the solute. On average, three models tested in the present work perform similarly. In particular, we achieved mean unsigned errors of 1.3 (SM8), 2.0 (SM8AD), and 2.6 kcal/mol (SMD) for the aqueous free energies of 30 out of 31 compounds with known reference data involved in Task 1 and mean unsigned errors of 2.7 (SM8), 1.8 (SM8AD), and 2.4 kcal/mol (SMD) in the free energy differences (tautomeric ratios) for 21 tautomeric pairs in aqueous solution involved in Task 2.

Protein-ion binding process on finite macromolecular concentration. A Poisson-Boltzmann and Monte Carlo study

Electrostatic interactions are one of the key driving forces for protein-ligands complexation. Different levels for the theoretical modeling of such processes are available on the literature. Most of the studies on the Molecular Biology field are performed within numerical solutions of the Poisson-Boltzmann Equation and the dielectric continuum models framework. In such dielectric continuum models, there are two pivotal questions: (a) how the protein dielectric medium should be modeled, and (b) what protocol should be used when solving this effective Hamiltonian. By means of Monte Carlo (MC) and Poisson-Boltzmann (PB) calculations, we define the applicability of the PB approach with linear and nonlinear responses for macromolecular electrostatic interactions in electrolyte solution, revealing some physical mechanisms and limitations behind it especially due the raise of both macromolecular charge and concentration out of the strong coupling regime. A discrepancy between PB and MC for binding constant shifts is shown and explained in terms of the manner PB approximates the excess chemical potentials of the ligand, and not as a consequence of the nonlinear thermal treatment and/or explicit ion-ion interactions as it could be argued. Our findings also show that the nonlinear PB predictions with a low dielectric response well reproduce the pK shifts calculations carried out with an uniform dielectric model. This confirms and completes previous results obtained by both MC and linear PB calculations.

Design of ligand binding to an engineered protein cavity using virtual screening and thermal up-shift evaluation

Proteins could be used to carry and deliver small compounds. As a tool for designing ligand binding sites in protein cores, a three-step virtual screening method is presented that has been optimised using existing data on T4 lysozyme complexes and tested in a newly engineered cavity in flavodoxin. The method can pinpoint, in large databases, ligands of specific protein cavities. In the first step, physico-chemical filters are used to screen the library and discard a majority of compounds. In the second step, a flexible, fast docking procedure is used to score and select a smaller number of compounds as potential binders. In the third step, a finer method is used to dock promising molecules of the hit list into the protein cavity, and an optimised free energy function allows discarding the few false positives by calculating the affinity of the modelled complexes. To demonstrate the portability of the method, several cavities have been designed and engineered in the flavodoxin from Anabaena PCC 7119, and the W66F/L44A double mutant has been selected as a suitable host protein. The NCI database has then been screened for potential binders, and the binding to the engineered cavity of five promising compounds and three tentative non-binders has been experimentally tested by thermal up-shift assays and spectroscopic titrations. The five tentative binders (some apolar and some polar), unlike the three tentative non-binders, are shown to bind to the host mutant and, importantly, not to bind to the wild type protein. The three-step virtual screening method developed can thus be used to identify ligands of buried protein cavities. We anticipate that the method could also be used, in a reverse manner, to identify natural or engineerable protein cavities for the hosting of ligands of interest.

Modelling and study of cyclosporin A and related compounds in complexes with a Trypanosoma cruzi cyclophilin

Cyclophilins (CyPs) are enzymes involved in protein folding, catalyzing the isomerisation of peptidyl prolyl bonds in proteins and peptides between the cis- and trans-conformations. They are also the major cellular target for the immunosuppressive drug Cyclosporin A (CsA). In Trypanosoma cruzi, the most abundantly expressed CyP is an isoform of 19 kDa, TcCyP19, in which the enzymatic activity is inhibited by CsA. Among a reported set of CsA analogues, two non-immunosuppressive compounds, H-7-94 and F-7-62, proved to be the best inhibitors of TcCyP19 enzymatic activity as well as the most efficient trypanocidal drugs. With the objective of analysing, at the molecular level, how the structural differences between the three above-mentioned inhibitors justify their different inhibitory activity on TcCyP19, three-dimensional molecular modelling structures were generated to computationally simulate behaviours and interactions. An energy-minimized model of each binary complex in water with ions was obtained. These models were then used as starting point for molecular dynamic simulations, performed with GROMOS96 program. With the resulting set of co-ordinates and energies, a comparison of the interaction between CsA and both CsA analogues in T. cruzi and human cyclophilins were performed. Within the different magnitudes analysed, the total potential complex energy exhibited the best correlation with the experimental data. The results obtained in this study support the use of this methodology when designing new lead inhibitor compounds.

Novel, Highly Potent Adenosine Deaminase Inhibitors Containing the Pyrazolo[3,4-d]pyrimidine Ring System. Synthesis, Structure−Activity Relationships, and Molecular Modeling Studies

This study reports the synthesis of a number of 1- and 2-alkyl derivatives of the 4-aminopyrazolo[3,4-d]pyrimidine (APP) nucleus and their evaluation as inhibitors of ADA from bovine spleen. The 2-substituted aminopyrazolopyrimidines proved to be potent inhibitors, most of them exhibiting K(i) values in the nanomolar/subnanomolar range. In this series the inhibitory activity is enhanced with the increase in length of the alkyl chain, reaching a maximum with the n-decyl substituent. Insertion of a 2'-hydroxy group in the n-decyl chain gave 3k, whose (R)-isomer displayed the highest inhibitory potency of the series (K(i) 0.053 nM), showing an activity 2 orders of magnitude higher than that of (+)-EHNA (K(i) 1.14 nM), which was taken as the reference standard. Docking simulations of aminopyrazolopyrimidines into the ADA binding site were also performed, to rationalize the structure-activity relationships of this class of inhibitors.

Computer-aided organic synthesis

It is tempting for those in the field of organic synthesis to liken the process of retrosynthesis to a game of chess. That the world chess champion was recently defeated by a computer leads us to think that perhaps new and powerful computing methods could be applied to synthetic problems. Here the analogy between synthesis and chess is outlined. Achievements in the 35-year history of computer-aided synthetic design are described, followed by some more recent developments.

Structural Consequences of D481N/K483Q Mutation at Glycine Binding Site of NMDA Ionotropic Glutamate Receptors: A Molecular Dynamics Study

N-Methyl-D-Aspartate (NMDA) receptors are the ligand gated as well as voltage sensitive ionotropic glutamate receptors, widely distributed in the vertebrate central nervous system and they play critical role in the pathogenesis of schizophrenia. Molecular dynamics simulations have been carried out on high resolution crystal structure of NR1 subunit of NMDA receptor ligand binding core (S1S2) in four different conformations. We have investigated consequence of D481N/K483Q double mutation of NR1 subunit from simulation results of (a) glycine bound form (WG), (b) unbound (closed-apo) form (WOG), (c) a double mutated form (DM), and (d) the antagonist (5,7-dichlorokynuric acid) bound form (DCKA). The MD simulations and simulated annealing for 4ns show a distinct conformation for the double mutated conformation that neither follows the antagonist nor apo conformation. There are two distinct sites, loop1 and loop2 where the double mutated structure in its glycine bound form shows significant RMSD deviations as compared to the wild-type. The interactions of glycine with the receptor remain theoretically unchanged in the double mutated structure and there is no detachment of S1S2 domains. The results suggest that separation of S1 and S2 domains may not be essential for channel inactivation. Therefore, it is hypothesized that hypoactivation of NMDA receptor channels may arise out of the conformational changes at non-conserved Loop1 and Loop2 regions observed in the mutated structure. The Loop1 and Loop2 regions responsible for inter-subunit interactions in a functional NMDA receptor, may therefore, render the ligand bound form defunct. This may account for behavioral anomalies due to receptor inactivation seen in grin1 mutated mice.

Prediction of HIV-1 Integrase/Viral DNA Interactions in the Catalytic Domain by Fast Molecular Docking

This study details the separate analyses of binding specificity of HIV-1 integrase (IN) and viral B-DNA forms through ligand-receptor docking studies by means of a fast molecular docking method. The application of solvated electrostatics with the University of Houston Brownian Dynamics Program (UHBD) and configurational sampling by the Daughter of Turnip (DOT) docking program resulted in the computation of energies of more than 113 billion configurations for each ligand-receptor docking study, a procedure considered computationally intractable a few years ago. A specific binding pattern of viral DNA to the IN catalytic domain region has been predicted as a result of these calculations. In a representative docked configuration, we observe the 3'-hydroxyl of the conserved deoxyadenosine to be close to one of the two divalent metal ions that are necessary for catalysis. A superimposition of our energy-minimized docked complex on representative structures from a molecular dynamics (MD) simulation of a crystallographically resolved IN/inhibitor complex revealed an overlap of viral DNA with the inhibitor, indicating that the bound inhibitor might operate by blocking substrate binding. The DOT docking calculation also identified a second, adjacent DNA-binding site, which we believe is the nonspecific host DNA binding site. The binding pattern predicted by DOT complements previous electrostatics, MD simulation, photo-cross-linking, and mutagenesis studies and also provides a further refinement of the IN/viral DNA binding interaction as a basis for new structure-based design efforts.

Predicting protein–ligand binding affinities: a low scoring game?

We have investigated the performance of five well known scoring functions in predicting the binding affinities of a diverse set of 205 protein-ligand complexes with known experimental binding constants, and also on subsets of mutually similar complexes. We have found that the overall performance of the scoring functions on the diverse set is disappointing, with none of the functions achieving r(2) values above 0.32 on the whole dataset. Performance on the subsets was mixed, with four of the five functions predicting fairly well the binding affinities of 35 proteinases, but none of the functions producing any useful correlation on a set of 38 aspartic proteinases. We consider two algorithms for producing consensus scoring functions, one based on a linear combination of scores from the five individual functions and the other on averaging the rankings produced by the five functions. We find that both algorithms produce consensus functions that generally perform slightly better than the best individual scoring function on a given dataset.

Binding mode prediction for a flexible ligand in a flexible pocket using multi-conformation simulated annealing pseudo crystallographic refinement

We describe multi-conformation simulated annealing-pseudo-crystallographic refinement (MCSA-PCR), a technique developed for predicting the binding mode of a flexible ligand in a flexible binding pocket. To circumvent the local-minimum problem efficiently, this method performs multiple independent cycles of simulated annealing with explicit solvent, "growing" the ligand in the binding pocket each time. From the ensemble of structures, a pseudo-crystallographic electron density map is calculated, and then conventional crystallographic refinement methods are used to best fit a single, optimal structure into the density map. The advantage of the MCSA-PCR method is that it provides a direct means to evaluate the accuracy and uniqueness of the calculated solution, provides a measure of ligand and protein dynamics from the refined B-factors, and facilitates comparison with X-ray crystallographic data. Here, we show that our MCSA-PCR method succeeds in predicting the correct binding mode of the VSV8 peptide to the major histocompatibility complex (MHC) receptor. Importantly, there is a significant correlation between the experimentally determined crystallographic water molecules and water density observed in the pseudo map by MCSA-PCR. Furthermore, comparison of different approaches for extracting a single, most probable structure from the calculated ensemble reveals the power of the PCR method and provides insights into the nature of the energetic landscape.

Conformational dynamics of subtilisin-chymotrypsin inhibitor 2 complex by coarse-grained simulations

An off-lattice dynamic Monte Carlo (MC) method is used to investigate the conformational dynamics of chymotrypsin inhibitor 2 (CI2) and subtilisin in both free and complex forms over two time windows, referring to short and long time scales. The conformational dynamics of backbone bonds analysed from several independent trajectories reveal that: Both the inhibitor and the enzyme are restricted in their bond rotations, excluding a few bonds, upon binding; the effect being greatest for the loop regions, and for the inhibitor. A cooperativity in the near-neighbor bond rotations are observed on both time scales, whereas the cooperative rotations of the bonds far along the sequence appear only in the long time window, and the latter time window is where most of the interactions between the inhibitor and the enzyme are observed. Upon binding, the cooperatively rotating parts of the inhibitor and the enzyme are readjusted compared to their free forms, and new correlations appear. The binding loop, although it is the closest contact region, is not the only part of the inhibitor involved in the interactions with the enzyme. Loops 3 and 8 and the helices F and G in bound enzyme and the binding loop of the inhibitor contribute at the most to the collective motions of whole structure on the slow time scale and are apparently important for enzyme-inhibitor interactions and function. The results in general provide evidence for the contribution of the loops with cooperative motions to the extensive communication network of the complex.

Active site binding modes of HIV-1 integrase inhibitors

Using the crystal structure of the first complex of the HIV-1 integrase catalytic core domain with an inhibitor bound to the active site, structural models for the interaction of various inhibitors with integrase were generated by computational docking. For the compound of the crystallographic study, binding modes unaffected by crystal packing have recently been proposed. Although a large search region was used for the docking simulations, the ligands investigated here are found to bind preferably in similar ways close to the active site. The binding site is formed by residues 64-67, 116, 148, 151-152, 155-156, and 159, as well as by residue 92 in case of the largest ligand of the series. The coherent picture of possible interactions of small-molecule inhibitors at the active site provides an improved basis for structure-based ligand design. The recurring motif of tight interaction with the two lysine residues 156 and 159 is suggested to be of prime importance.

A fast method to predict protein interaction sites from sequences

A simple method for predicting residues involved in protein interaction sites is proposed. In the absence of any structural report, the procedure identifies linear stretches of sequences as "receptor-binding domains" (RBDs) by analysing hydrophobicity distribution. The sequences of two databases of non-homologous interaction sites eliciting various biological activities were tested; 59-80 % were detected as RBDs. A statistical analysis of amino acid frequencies was carried out in known interaction sites and in predicted RBDs. RBDs were predicted from the 80,000 sequences of the Swissprot database. In both cases, arginine is the most frequently occurring residue. The RBD procedure can also detect residues involved in specific interaction sites such as the DNA-binding (95 % detected) and Ca-binding domains (83 % detected). We report two recent analyses; from the prediction of RBDs in sequences to the experimental demonstration of the functional activities. The examples concern a retroviral Gag protein and a penicillin-binding protein. We support that this method is a quick way to predict protein interaction sites from sequences and is helpful for guiding experiments such as site-specific mutageneses, two-hybrid systems or the synthesis of inhibitors.

Structure/thermodynamics relationships of lectin-saccharide complexes: the Erythrina corallodendron case

Molecular dynamics (MD) simulations of Erythrina corallodendron lectin binding to a monosaccharide, alpha-galactose, and a disaccharide, N-acetyl lactosamine, have been performed in order to investigate the relationship between structure and thermodynamics. A simulated annealing protocol has been used to generate ensembles of structures for the two complexes, from which both qualitative and quantitative information on binding dynamics have been extracted. The ensembled averaged lectin-saccharide interaction enthalpy is equivalent for both sugars, whereas the calculation based on the X-ray structures does show a difference. Within large statistical errors, the calculated 'binding enthalpy' is also the same for the two systems. These errors arise largely from terms involving solvent and are a typical limitation of current MD simulations. Significant qualitative differences in binding between the two complexes are, however, observed over the ensembles. These could be important for unraveling the structure/thermodynamic relationship. Stated simply, there are a greater number of binding options available to the disaccharide compared to the monosaccharide. The implications of alternative binding states on thermodynamic parameters and the 'breaking of enthalpy-entropy compensation' are discussed. The role of solvent in lectin-saccharide complex formation is suggested to be significant.

Automated docking of ligands to antibodies: methods and applications

Many approaches to studying protein-ligand interactions by computational docking are currently available. Given the structures of a protein and a ligand, the ultimate goal of all docking methods is to predict the structure of the resulting complex. This requires a suitable representation of molecular structures and properties, search algorithms to efficiently scan the configuration space for favorable interaction geometries, and accurate scoring functions to evaluate and rank the generated orientations. For many of the available methods, tests on experimentally known antibody-antigen or antibody-hapten complexes have appeared in the literature. In addition, some of them have been used in predictive studies on antibody-ligand interactions to provide structural insights where adequate experimental information is missing. The AutoDock program is presented as example of a method for flexibly docking ligands to antibodies. Applying parameters of the second-generation AMBER force field, three antibody-hapten complexes (AN02, DB3, NC6.8) are used as new test cases to analyze the ability of the method to reproduce experimental findings. The X-ray structures could be reconstituted and the corresponding solutions were ranked with best energy score in all cases. Docking to the free instead of the complexed NC6.8 structure indicated the limits of the rigid protein treatment, although fairly good guesses about the location of the binding site and the contact residues could still be obtained if conformational flexibility was allowed at least in the ligand.

Computer simulation of the interaction of non-steroidal anti-inflammatory drugs: indoprofen and NS398 with cyclooxygenase

We have applied computer simulation technique to study interaction of two anti-inflammatory drugs (NSAIDs) indoprofen and NS398 with cyclooxygenase (COX-1 and COX-2) enzymes. We have also investigated conformational flexibility of the two drugs by systematic search and simulated annealing molecular dynamics (SAMD) methods. Both the drugs were docked in the cyclooxygenase channel using in house docking program IMF1. The complexes were energy minimised by molecular mechanics (MM) method. These were heated for 30 picoseconds (ps), equilibrated for 110 ps at 300K and subjected to 'production simulation' for 110 ps by molecular dynamics (MD) method using Sanderís module of AMBER 5.0 package and united atom force field mostly from PARM96.DAT. Integration was carried out with time step of 0.001 ps, distance dependent di-electric constant with scaling factor 2.0 for 1-4 interaction and cut-off distance for non-bonded pair-list equal to 8A. The non-bonded pair-list was upgraded after every 20 cycles. The coordinate output from MD trajectories is analysed using analysis package of AMBER 5.0, MOLMOL, P-CURVES 3.0 and in house packages: ANALMD, ANALP1. We have observed perturbative changes in COX-1 and COX-2 structures due to indoprofen and NS398. In case of indoprofen specific changes between COX-1 and COX-2 were noted in helix D, H6, S6 and helix H8 in the cyclooxygenase cavity. In case of NS398 these were in helix B in membrane binding domain, helix H6, S8 and S10 in cyclooxygenase cavity and helices H14-H16 in small lobe close to haem binding region. Implications of these results in enzyme selectivity by NSAIDs is discussed here.

Chance favors the prepared mind--from serendipity to rational drug design

Accidental discoveries always played an important role in science, especially in the search for new drugs. Several examples of serendipitous findings, leading to therapeutically useful drugs, are presented and discussed. Captopril, an antihypertensive Angiotensin-converting enzyme inhibitor, was the first drug that could be derived from a structural model of a protein. Dorzolamide, a Carboanhydrase inhibitor for the treatment of glaucoma, and the HIV protease inhibitors Saquinavir, Indinavir, Ritonavir, and Nelfinavir are further examples of therapeutically used drugs from structure-based design. More enzyme inhibitors, e.g. the anti-influenza drugs Zanamivir and GS 4104, are in clinical development. In the absence of a protein 3D structure, the 3D structures of certain ligands may be used for rational design. This approach is exemplified by the design of specifically acting integrin receptor antagonists. In the last years, combinatorial and computational approaches became important methods for rational drug design. SAR by NMR searches for low-affinity ligands that bind to proximal subsites of an enzyme; linkage with an appropriate tether produces nanomolar inhibitors. The de novo design program LUDI and the docking program FlexX are tools for the computer-aided design of protein ligands. Work is in progress to combine such approaches to strategies for combinatorial drug design.

Identification and energetic ranking of possible docking sites for pterin on dihydrofolate reductase

The reliability of new methodology for detecting sites for ligand binding on the surfaces of proteins has been tested using a range of dihydrofolate reductase (DHFR) crystal structures. Docking of the pterin molecule to ten such DHFR structures has been examined. Initial docking sites were selected using the VDW-FFT method we have developed recently. This procedure was followed by rigid geometry optimization and solvation energy calculations using our parametrized reaction field multipoles (PRFM) method and the finite difference solution of the Poisson equation (FDPB) method. Two different sets of MM parameters, from the OPLS and Amber94 force fields, have been used. In eight cases the energy of the complexes with pterin bound at the active site was the lowest with the recent Amber94 parameters. In one case the spurious first-ranked site was only 1.8 kcal/mol lower in energy compared with the active site. The other 'failure' of the method may, in fact, represent a valid initial binding site. The calculations with the old OPLS parameters gave slightly worse results.

Modelling a 3D structure for EgDf1 from Echinococcus granulosus: putative epitopes, phosphorylation motifs and ligand

EgDf1 is a developmentally regulated protein from the parasite Echinococcus granulosus related to a family of hydrophobic ligand binding proteins. This protein could play a crucial role during the parasite life cycle development since this organism is unable to synthetize most of their own lipids de novo. Furthermore, it has been shown that two related protein from other parasitic platyhelminths (Fh15 from Fasciola hepatica and Sm14 from Schistosoma mansoni) are able to confer protective inmunity against experimental infection in animal models. A three-dimensional structure would help establishing structure/function relationships on a knowledge based manner. 3D structures for EgDf1 protein were modelled by using myelin P2 (mP2) and intestine fatty acid binding protein (I-FABP) as templates. Molecular dynamics techniques were used to validate the models. Template mP2 yielded the best 3D structure for EgDf1. Palmitic and oleic acids were docked inside EgDf1. The present theoretical results suggest definite location in the secondary structure of the epitopic regions, consensus phosphorylation motifs and oleic acid as a good ligand candidate to EgDf1. This protein might well be involved in the process of supplying hydrophobic metabolites for membrane biosynthesis and for signaling pathways.

Predictive docking of protein-protein and protein-DNA complexes

Recent developments in algorithms to predict the docking of two proteins have considered both the initial rigid-body global search and subsequent screening and refinement. The result of two blind trials of protein docking are encouraging--for complexes that are not too large and do not undergo sizeable conformational change upon association, the algorithms are now able to suggest reasonably accurate models.

LIGSITE: automatic and efficient detection of potential small molecule-binding sites in proteins

LIGSITE is a new program for the automatic and time-efficient detection of pockets on the surface of proteins that may act as binding sites for small molecule ligands. Pockets are identified with a series of simple operations on a cubic grid. Using a set of receptor-ligand complexes we show that LIGSITE is able to identify the binding sites of small molecule ligands with high precision. The main advantage of LIGSITE is its speed. Typical search times are in the range of 5 to 20 s for medium-sized proteins. LIGSITE is therefore well suited for identification of pockets in large sets of proteins (e.g., protein families) for comparative studies. For graphical display LIGSITE produces VRML representations of the protein-ligand complex and the binding site for display with a VRML viewer such as WebSpace from SGI.

Can drugs be designed?

Many examples are now emerging of the successful use of rational, structure-based methods in drug discovery. Of particular note is the development of imaginative NMR-based methods for rapid routes to ligand design. Our understanding of the chemistry underlying protein-ligand interactions, however, remains relatively poor and a major limitation in our ability to truly design drugs.

Prediction of protein side-chain rotamers from a backbone-dependent rotamer library: a new homology modeling tool

Modeling by homology is the most accurate computational method for translating an amino acid sequence into a protein structure. Homology modeling can be divided into two sub-problems, placing the polypeptide backbone and adding side-chains. We present a method for rapidly predicting the conformations of protein side-chains, starting from main-chain coordinates alone. The method involves using fewer than ten rotamers per residue from a backbone-dependent rotamer library and a search to remove steric conflicts. The method is initially tested on 299 high resolution crystal structures by rebuilding side-chains onto the experimentally determined backbone structures. A total of 77% of chi1 and 66% of chi(1 + 2) dihedral angles are predicted within 40 degrees of their crystal structure values. We then tested the method on the entire database of known structures in the Protein Data Bank. The predictive accuracy of the algorithm was strongly correlated with the resolution of the structures. In an effort to simulate a realistic homology modeling problem, 9424 homology models were created using three different modeling strategies. For prediction purposes, pairs of structures were identified which shared between 30% and 90% sequence identity. One strategy results in 82% of chi1 and 72% chi(1 + 2) dihedral angles predicted within 40 degrees of the target crystal structure values, suggesting that movements of the backbone associated with this degree of sequence identity are not large enough to disrupt the predictive ability of our method for non-native backbones. These results compared favorably with existing methods over a comprehensive data set.

Structure prediction: folding proteins by pattern recognition

Although we are still a long way from being able to predict the details of protein structure from the underlying chemistry, slow but steady progress is being made at modeling structural features by recognizing the patterns that connect sequence to structure.