Comparative Analysis of Protein-Bound Ligand Conformations with Respect to Catalyst's Conformational Space Subsampling Algorithms

TFD: Torsion Fingerprints as a new measure to compare small molecule conformations

Advantages like intuitive interpretation, objectivity, general applicability, and its easy, automated calculation make the rmsd (root-mean-squared deviation) the measure of choice for the investigation of the accuracy of conformational model generators. For comparing conformations of a single molecule this is a clearly superior method. Single molecule analysis is, however, a rare scenario. Typically, conformations are generated for huge corporate or external vendor databases of high diversity which are then further investigated with high-throughput computational methods like docking or pharmacophore searching, in virtual screening campaigns. Representative subsets for accuracy investigations of computational methods need to mimic this diversity. Averaged rmsd values over these data sets are frequently used to assess the accuracy of the methods. There are, however, significant weaknesses in rmsd comparisons for such kind of data sets. The interpretation is for example no longer intuitive because what can be expected in terms of good or bad rmsd values crucially depends on the data set composition like size or number of rotatable bonds of the underlying molecules. Further, rmsd lacks normalization which might result in very high averaged rmsd values for highly flexible molecules and thus might completely skew results. We have developed a novel measure to compare conformations of molecules called Torsion Fingerprint Deviation (TFD). It extracts, weights, and compares Torsion Fingerprints from a query molecule and generated conformations under consideration of acyclic bonds as well as ring systems. TFD is alignment-free and overcomes major limitations of rmsd while retaining its advantages.

Successful prediction of substrate-binding pocket in SLC17 transporter sialin

Secondary active transporters from the SLC17 protein family are required for excitatory and purinergic synaptic transmission, sialic acid metabolism, and renal function, and several members are associated with inherited neurological or metabolic diseases. However, molecular tools to investigate their function or correct their genetic defects are limited or absent. Using structure-activity, homology modeling, molecular docking, and mutagenesis studies, we have located the substrate-binding site of sialin (SLC17A5), a lysosomal sialic acid exporter also recently implicated in exocytotic release of aspartate. Human sialin is defective in two inherited sialic acid storage diseases and is responsible for metabolic incorporation of the dietary nonhuman sialic acid N-glycolylneuraminic acid. We built cytosol-open and lumen-open three-dimensional models of sialin based on weak, but significant, sequence similarity with the glycerol-3-phosphate and fucose permeases from Escherichia coli, respectively. Molecular docking of 31 synthetic sialic acid analogues to both models was consistent with inhibition studies. Narrowing the sialic acid-binding site in the cytosol-open state by two phenylalanine to tyrosine mutations abrogated recognition of the most active analogue without impairing neuraminic acid transport. Moreover, a pilot virtual high-throughput screening of the cytosol-open model could identify a pseudopeptide competitive inhibitor showing >100-fold higher affinity than the natural substrate. This validated model of human sialin and sialin-guided models of other SLC17 transporters should pave the way for the identification of inhibitors, glycoengineering tools, pharmacological chaperones, and fluorescent false neurotransmitters targeted to these proteins.

Mangiferin, an anti-HIV-1 agent targeting protease and effective against resistant strains

The anti-HIV-1 activity of mangiferin was evaluated. Mangiferin can inhibit HIV-1_ⅢB induced syncytium formation at non-cytotoxic concentrations, with a 50% effective concentration (EC₅₀) at 16.90 μM and a therapeutic index (TI) above 140. Mangiferin also showed good activities in other laboratory-derived strains, clinically isolated strains and resistant HIV-1 strains. Mechanism studies revealed that mangiferin might inhibit the HIV-1 protease, but is still effective against HIV peptidic protease inhibitor resistant strains. A combination of docking and pharmacophore methods clarified possible binding modes of mangiferin in the HIV-1 protease. The pharmacophore model of mangiferin consists of two hydrogen bond donors and two hydrogen bond acceptors. Compared to pharmacophore features found in commercially available drugs, three pharmacophoric elements matched well and one novel pharmacophore element was observed. Moreover, molecular docking analysis demonstrated that the pharmacophoric elements play important roles in binding HIV-1 protease. Mangiferin is a novel nonpeptidic protease inhibitor with an original structure that represents an effective drug development strategy for combating drug resistance.

Computational ligand-based rational design: Role of conformational sampling and force fields in model development

A significant number of drug discovery efforts are based on natural products or high throughput screens from which compounds showing potential therapeutic effects are identified without knowledge of the target molecule or its 3D structure. In such cases computational ligand-based drug design (LBDD) can accelerate the drug discovery processes. LBDD is a general approach to elucidate the relationship of a compound's structure and physicochemical attributes to its biological activity. The resulting structure-activity relationship (SAR) may then act as the basis for the prediction of compounds with improved biological attributes. LBDD methods range from pharmacophore models identifying essential features of ligands responsible for their activity, quantitative structure-activity relationships (QSAR) yielding quantitative estimates of activities based on physiochemical properties, and to similarity searching, which explores compounds with similar properties as well as various combinations of the above. A number of recent LBDD approaches involve the use of multiple conformations of the ligands being studied. One of the basic components to generate multiple conformations in LBDD is molecular mechanics (MM), which apply an empirical energy function to relate conformation to energies and forces. The collection of conformations for ligands is then combined with functional data using methods ranging from regression analysis to neural networks, from which the SAR is determined. Accordingly, for effective application of LBDD for SAR determinations it is important that the compounds be accurately modelled such that the appropriate range of conformations accessible to the ligands is identified. Such accurate modelling is largely based on use of the appropriate empirical force field for the molecules being investigated and the approaches used to generate the conformations. The present chapter includes a brief overview of currently used SAR methods in LBDD followed by a more detailed presentation of issues and limitations associated with empirical energy functions and conformational sampling methods.

Computer- and structure-based lead design for epigenetic targets

The term epigenetics is defined as inheritable changes that influence the outcome of a phenotype without changes in the genome. Epigenetics is based upon DNA methylation and posttranslational histone modifications. While there is much known about reversible acetylation as a posttranslational modification, research on reversible histone methylation is still emerging, especially with regard to drug discovery. As aberrant epigenetic modifications have been linked to many diseases, inhibitors of histone modifying enzymes are very much in demand. This article will summarize the progress on small molecule epigenetic inhibitors identified by structure- and computer-based approaches.

Pharmacophore-based virtual screening

This chapter is a review of the most recent developments in the field of pharmacophore modeling, covering both methodology and application. Pharmacophore-based virtual screening is nowadays a mature technology, very well accepted in the medicinal chemistry laboratory. Nevertheless, like any empirical approach, it has specific limitations and efforts to improve the methodology are still ongoing. Fundamentally, the core idea of "stripping" functional groups of their actual chemical nature in order to classify them into very few pharmacophore types, according to their dominant physico-chemical features, is both the main advantage and the main drawback of pharmacophore modeling. The advantage is the one of simplicity - the complex nature of noncovalent ligand binding interactions is rendered intuitive and comprehensible by the human mind. Although computers are much better suited for comparisons of pharmacophore patterns, a chemist's intuition is primarily scaffold-oriented. Its underlying simplifications render pharmacophore modeling unable to provide perfect predictions of ligand binding propensities - not even if all its subsisting technical problems would be solved. Each step in pharmacophore modeling and exploitation has specific drawbacks: from insufficient or inaccurate conformational sampling to ambiguities in pharmacophore typing (mainly due to uncertainty regarding the tautomeric/protonation status of compounds), to computer time limitations in complex molecular overlay calculations, and to the choice of inappropriate anchoring points in active sites when ligand cocrystals structures are not available. Yet, imperfections notwithstanding, the approach is accurate enough in order to be practically useful and actually is the most used virtual screening technique in medicinal chemistry - notably for "scaffold hopping" approaches, allowing the discovery of new chemical classes carriers of a desired biological activity.

Flavonoids as protein kinase inhibitors for cancer chemoprevention: direct binding and molecular modeling

Protein kinases play crucial roles in the regulation of multiple cell signaling pathways and cellular functions. Deregulation of protein kinase function has been implicated in carcinogenesis. The inhibition of protein kinases has emerged as an important target for cancer chemoprevention and therapy. Accumulated data revealed that flavonoids exert chemopreventive effects through acting at protein kinase signaling pathways, more than as conventional hydrogen-donating antioxidants. Recent studies show that flavonoids can bind directly to some protein kinases, including Akt/protein kinase B (Akt/PKB), Fyn, Janus kinase 1 (JAK1), mitogen-activated protein kinase kinase 1 (MEK1), phosphoinositide 3-kinase (PI3K), mitogen-activated protein (MAP) kinase kinase 4 (MKK4), Raf1, and zeta chain-associated 70-kDa protein (ZAP-70) kinase, and then alter their phosphorylation state to regulate multiple cell signaling pathways in carcinogenesis processes. In this review, we report recent results on the interactions of flavonoids and protein kinases, especially their direct binding and molecular modeling. The data suggest that flavonoids act as protein kinase inhibitors for cancer chemoprevention that were thought previously as conventional hydrogen-donating antioxidant. Moreover, the molecular modeling data show some hints for creating natural compound-based protein kinase inhibitors for cancer chemoprevention and therapy.

Drug-like bioactive structures and conformational coverage with the LigPrep/ConfGen suite: comparison to programs MOE and catalyst

Computational conformational sampling underpins many aspects of small molecule modeling and design in pharmaceutical work. This work examined in detail the widely distributed LigPrep/ConfGen software suite and the conformational models it produces for drug-like compounds. We also compare LigPrep/ConfGen to MOE and Catalyst. Tests of the conformational sampling protocols included the reproduction of known bioactive structures of ligands, characterization of the size, coverage and diversity of the output conformational models, and relative computation times. The present tests will help the user to make informed choices among the predefined ConfGen protocols (Very fast, Fast, Intermediate, and Comprehensive), and the adjustable input parameters. The parameters governing the initial compound preparation (LigPrep) and the subsequent conformational sampling were explored. This analysis has led to a new protocol called "ConfGen Optimized", which improves upon the predefined protocols. ConfGen Optimized is computationally tractable and reproduced 80% of the bioactive structures within 1 A, versus 66% for the default ConfGen Fast protocol. We also addressed the issue of the reproduction of compact/folded bioactive structures by ConfGen. It involved the compilation of a new set of 50 folded diverse drug-like bioactive structures. This indicates that heuristics penalizing folded conformers hinder reproduction of some binding modes. Overall, ConfGen offers great flexibility of use and provides a valuable addition to the molecular modeling toolbox.

ConfGen: a conformational search method for efficient generation of bioactive conformers

We describe the methodology, parametrization, and application of a conformational search method, called ConfGen, designed to efficiently generate bioactive conformers. We define efficiency as the ability to generate a bioactive conformation within a small total number of conformations using a reasonable amount of computer time. The method combines physics-based force field calculations with empirically derived heuristics designed to achieve efficient searching and prioritization of the ligand's conformational space. While many parameter settings are supported, four modes spanning a range of speed and quality trades-offs are defined and characterized. The validation set used to test the method is composed of ligands from 667 crystal structures covering a broad array of target and ligand classes. With the fastest mode, ConfGen uses an average of 0.5 s per ligand and generates only 14.3 conformers per ligand, at least one of which lies within 2.0 A root-mean-squared deviation of the crystal structure for 96% of the ligands. The most computationally intensive mode raises this recovery rate to 99%, while taking 8 s per ligand. Combining multiple search modes to "fill-in" holes in the conformation space or energy minimizing using an all-atom force field each lead to improvements in the recovery rates at higher resolutions. Overall, ConfGen is at least as good as competing programs at high resolution and demonstrates higher efficiency at resolutions sufficient for many downstream applications, such as pharmacophore modeling.

Docking-enabled pharmacophore model for histone deacetylase 8 inhibitors and its application in anti-cancer drug discovery

Zinc-dependent histone deacetylase 8 removes the epsilon-acetyl groups present in the N-terminal lysine residues of histone proteins, thereby restricting various transcription factors from being expressed. Inhibition of this enzyme has been reported to be a novel strategy in cancer treatment. To identify novel and diverse leads for use in potent histone deacetylase 8 inhibitor design, a pharmacophore model showing high correlation between experimental and estimated activities was generated using the best conformations of training set compounds from molecular docking experiments. The best pharmacophore model was validated using four different strategies and then used in database screening for novel virtual leads. Hit compounds were selected and subjected to molecular docking using GOLD. The top-scored compound was further optimized for improved binding. The optimization step led to a new set of compounds with both improved binding at the active site and estimated activities. The identified virtual leads could be used for designing potent histone deacetylase 8 inhibitors as anti-cancer therapeutics.

Dynamic clustering threshold reduces conformer ensemble size while maintaining a biologically relevant ensemble

Representing the 3D structures of ligands in virtual screenings via multi-conformer ensembles can be computationally intensive, especially for compounds with a large number of rotatable bonds. Thus, reducing the size of multi-conformer databases and the number of query conformers, while simultaneously reproducing the bioactive conformer with good accuracy, is of crucial interest. While clustering and RMSD filtering methods are employed in existing conformer generators, the novelty of this work is the inclusion of a clustering scheme (NMRCLUST) that does not require a user-defined cut-off value. This algorithm simultaneously optimizes the number and the average spread of the clusters. Here we describe and test four inter-dependent approaches for selecting computer-generated conformers, namely: OMEGA, NMRCLUST, RMS filtering and averaged-RMS filtering. The bioactive conformations of 65 selected ligands were extracted from the corresponding protein:ligand complexes from the Protein Data Bank, including eight ligands that adopted dissimilar bound conformations within different receptors. We show that NMRCLUST can be employed to further filter OMEGA-generated conformers while maintaining biological relevance of the ensemble. It was observed that NMRCLUST (containing on average 10 times fewer conformers per compound) performed nearly as well as OMEGA, and both outperformed RMS filtering and averaged-RMS filtering in terms of identifying the bioactive conformations with excellent and good matches (0.5 < RMSD < 1.0 A). Furthermore, we propose thresholds for OMEGA root-mean square filtering depending on the number of rotors in a compound: 0.8, 1.0 and 1.4 for structures with low (1-4), medium (5-9) and high (10-15) numbers of rotatable bonds, respectively. The protocol employed is general and can be applied to reduce the number of conformers in multi-conformer compound collections and alleviate the complexity of downstream data processing in virtual screening experiments.

DG-AMMOS: a new tool to generate 3d conformation of small molecules using distance geometry and automated molecular mechanics optimization for in silico screening

BACKGROUND

Discovery of new bioactive molecules that could enter drug discovery programs or that could serve as chemical probes is a very complex and costly endeavor. Structure-based and ligand-based in silico screening approaches are nowadays extensively used to complement experimental screening approaches in order to increase the effectiveness of the process and facilitating the screening of thousands or millions of small molecules against a biomolecular target. Both in silico screening methods require as input a suitable chemical compound collection and most often the 3D structure of the small molecules has to be generated since compounds are usually delivered in 1D SMILES, CANSMILES or in 2D SDF formats.

RESULTS

Here, we describe the new open source program DG-AMMOS which allows the generation of the 3D conformation of small molecules using Distance Geometry and their energy minimization via Automated Molecular Mechanics Optimization. The program is validated on the Astex dataset, the ChemBridge Diversity database and on a number of small molecules with known crystal structures extracted from the Cambridge Structural Database. A comparison with the free program Balloon and the well-known commercial program Omega generating the 3D of small molecules is carried out. The results show that the new free program DG-AMMOS is a very efficient 3D structure generator engine.

CONCLUSION

DG-AMMOS provides fast, automated and reliable access to the generation of 3D conformation of small molecules and facilitates the preparation of a compound collection prior to high-throughput virtual screening computations. The validation of DG-AMMOS on several different datasets proves that generated structures are generally of equal quality or sometimes better than structures obtained by other tested methods.

Novel method for the evaluation of 3D conformation generators

Conformation generation is a common and key process of computer-aided drug design. The reliability of the docking simulations, pharmacophore development, and 3D-QSAR analyses depends on the accuracy of conformations of small molecules used as input information for each program. Many conformation generators have been developed with the aim of efficiently generating all the putative bound conformations that small molecules adopt when they interact with macromolecules. Conformation generators have been evaluated by whether they can reproduce the experimentally determined bioactive conformations of bound small molecules. These bioactive conformations are usually obtained from publicly available crystal structures of protein-ligand complexes. However, it is difficult to obtain 2 or more than 2 bioactive conformations of one compound because multiple complex structures of a single molecule with various macromolecules are rarely available. Present methods, therefore, simply check whether a set of generated conformations includes the corresponding bioactive conformation. The overall validity of the entire set of generated conformations against bioactive conformation space has never been checked. In this work we developed a novel method for the evaluation of conformation generators, which makes it possible to measure the performance of a conformation generator based on its ability to reproduce the overall bioactive conformation space. We also determined the optimum parameter sets for OMEGA (OpenEye) based on the coverage of bioactive conformation space and computational efficiency. Our evaluation method elucidated that increasing the number of generated conformations is not necessary to obtain better reproducibility of the overall bioactive conformation space. Our method can be applied to the evaluation of the algorithm and/or design of the conformation generator program itself.

tCONCOORD-GUI: visually supported conformational sampling of bioactive molecules

Conformational flexibility of bioactive molecules poses a major challenge to computational biology. tCONCOORD generates structure ensembles based on geometrical considerations and has been successfully applied to predict protein conformational flexibility and essential degrees of freedom. We have now developed a graphical user interface (GUI) for tCONCOORD, which substantially facilitates the simulation setup and provides valuable insights into the structure analysis and constraint definition process in tCONCOORD. Moreover, users can influence the constraint definition process by interactively turning interactions on and off, defining completely rigid or flexible regions, or by applying artifical constraints that cause a biased sampling of the conformational space. This interface offers a versatile environment not only for the setup and analysis of tCONCOORD simulations, but also for molecular modeling and structure analysis in general. Both tCONCOORD* and the tCONCOORD-GUI(dagger) are distributed freely.

Cyndi: a multi-objective evolution algorithm based method for bioactive molecular conformational generation

BACKGROUND

Conformation generation is a ubiquitous problem in molecule modelling. Many applications require sampling the broad molecular conformational space or perceiving the bioactive conformers to ensure success. Numerous in silico methods have been proposed in an attempt to resolve the problem, ranging from deterministic to non-deterministic and systemic to stochastic ones. In this work, we described an efficient conformation sampling method named Cyndi, which is based on multi-objective evolution algorithm.

RESULTS

The conformational perturbation is subjected to evolutionary operation on the genome encoded with dihedral torsions. Various objectives are designated to render the generated Pareto optimal conformers to be energy-favoured as well as evenly scattered across the conformational space. An optional objective concerning the degree of molecular extension is added to achieve geometrically extended or compact conformations which have been observed to impact the molecular bioactivity (J Comput -Aided Mol Des 2002, 16: 105-112). Testing the performance of Cyndi against a test set consisting of 329 small molecules reveals an average minimum RMSD of 0.864 A to corresponding bioactive conformations, indicating Cyndi is highly competitive against other conformation generation methods. Meanwhile, the high-speed performance (0.49 +/- 0.18 seconds per molecule) renders Cyndi to be a practical toolkit for conformational database preparation and facilitates subsequent pharmacophore mapping or rigid docking. The copy of precompiled executable of Cyndi and the test set molecules in mol2 format are accessible in Additional file 1.

CONCLUSION

On the basis of MOEA algorithm, we present a new, highly efficient conformation generation method, Cyndi, and report the results of validation and performance studies comparing with other four methods. The results reveal that Cyndi is capable of generating geometrically diverse conformers and outperforms other four multiple conformer generators in the case of reproducing the bioactive conformations against 329 structures. The speed advantage indicates Cyndi is a powerful alternative method for extensive conformational sampling and large-scale conformer database preparation.

Toward accurate relative energy predictions of the bioactive conformation of drugs

Quantifying the relative energy of a ligand in its target-bound state (i.e. the bioactive conformation) is essential to understand the process of molecular recognition, to optimize the potency of bioactive molecules and to increase the accuracy of structure-based drug design methods. This is, nevertheless, seriously hampered by two interrelated issues, namely the difficulty in carrying out an exhaustive sampling of the conformational space and the shortcomings of the energy functions, usually based on parametric methods of limited accuracy. Matters are further complicated by the experimental uncertainty on the atomic coordinates, which precludes a univocal definition of the bioactive conformation. In this article we investigate the relative energy of bioactive conformations introducing two major improvements over previous studies: the use sophisticated QM-based methods to take into account both the internal energy of the ligand and the solvation effect, and the application of physically meaningful constraints to refine the bioactive conformation. On a set of 99 drug-like molecules, we find that, contrary to previous observations, two thirds of bioactive conformations lie within 0.5 kcal mol(-1) of a local minimum, with penalties above 2.0 kcal mol(-1) being generally attributable to structural determination inaccuracies. The methodology herein described opens the door to obtain quantitative estimates of the energy of bioactive conformations and can be used both as an aid in refining crystallographic structures and as a tool in drug discovery.

Fully automated molecular mechanics based induced fit protein-ligand docking method

We describe a method for docking a ligand into a protein receptor while allowing flexibility of the protein binding site. The method employs a multistep procedure that begins with the generation of protein and ligand conformations. An initial placement of the ligand is then performed by computing binding site hotspots. This initial placement is followed by a protein side-chain refinement stage that models protein flexibility. The final step of the process is an energy minimization of the ligand pose in the presence of the rigid receptor. Thus the algorithm models flexibility of the protein at two stages, before and after ligand placement. We validated this method by performing docking and cross docking studies of eight protein systems for which crystal structures were available for at least two bound ligands. The resulting rmsd values of the 21 docked protein-ligand complexes showed values of 2 A or less for all but one of the systems examined. The method has two critical benefits for high throughput virtual screening studies. First, no user intervention is required in the docking once the initial binding site selection has been made in the protein. Second, the initial protein conformation generation needs to be performed only once for a given binding region. Also, the method may be customized in various ways depending on the particular scenario in which dockings are being performed. Each of the individual steps of the method is fully independent making it straightforward to explore different variants of the high level workflow to further improve accuracy and performance.

Conformational sampling of druglike molecules with MOE and catalyst: implications for pharmacophore modeling and virtual screening

Computational conformational sampling is integral to small molecule pharmaceutical research, for detailed conformational analysis and high-throughput 3D library enumeration. These two regimes were tested in details for the general-purpose modeling program MOE, using its three conformational sampling methods, i.e. systematic search, stochastic search, and Conformation Import. The tests include i) identification of the global energy minimum, ii) reproduction of the bioactive conformation, iii) measures of conformational coverage with 3D descriptors, and iv) compute times. The bioactive conformers are from a new set of 256 diverse, druglike, protein-bound ligands compiled and analyzed with particular care. The MOE results are compared to those obtained from the established program Catalyst. Key parameters controlling the conformational coverage were varied systematically. Coverage and diversity of the conformational space were characterized with unique pharmacophore triplets or quadruplets. Overall, the protocols in both MOE and Catalyst performed well for their intended tasks. MOE performed at least as well as Catalyst for high-throughput library generation and detailed conformational modeling. This work provides a guide and specific recommendations regarding the usage of conformational sampling tools in MOE.

MS-DOCK: accurate multiple conformation generator and rigid docking protocol for multi-step virtual ligand screening

Background

The number of protein targets with a known or predicted tri-dimensional structure and of drug-like chemical compounds is growing rapidly and so is the need for new therapeutic compounds or chemical probes. Performing flexible structure-based virtual screening computations on thousands of targets with millions of molecules is intractable to most laboratories nor indeed desirable. Since shape complementarity is of primary importance for most protein-ligand interactions, we have developed a tool/protocol based on rigid-body docking to select compounds that fit well into binding sites.

Results

Here we present an efficient multiple conformation rigid-body docking approach, MS-DOCK, which is based on the program DOCK. This approach can be used as the first step of a multi-stage docking/scoring protocol. First, we developed and validated the Multiconf-DOCK tool that generates several conformers per input ligand. Then, each generated conformer (bioactives and 37970 decoys) was docked rigidly using DOCK6 with our optimized protocol into seven different receptor-binding sites. MS-DOCK was able to significantly reduce the size of the initial input library for all seven targets, thereby facilitating subsequent more CPU demanding flexible docking procedures.

Conclusion

MS-DOCK can be easily used for the generation of multi-conformer libraries and for shape-based filtering within a multi-step structure-based screening protocol in order to shorten computation times.

Evaluation of the performance of 3D virtual screening protocols: RMSD comparisons, enrichment assessments, and decoy selection--what can we learn from earlier mistakes?

Within the last few years a considerable amount of evaluative studies has been published that investigate the performance of 3D virtual screening approaches. Thereby, in particular assessments of protein-ligand docking are facing remarkable interest in the scientific community. However, comparing virtual screening approaches is a non-trivial task. Several publications, especially in the field of molecular docking, suffer from shortcomings that are likely to affect the significance of the results considerably. These quality issues often arise from poor study design, biasing, by using improper or inexpressive enrichment descriptors, and from errors in interpretation of the data output. In this review we analyze recent literature evaluating 3D virtual screening methods, with focus on molecular docking. We highlight problematic issues and provide guidelines on how to improve the quality of computational studies. Since 3D virtual screening protocols are in general assessed by their ability to discriminate between active and inactive compounds, we summarize the impact of the composition and preparation of test sets on the outcome of evaluations. Moreover, we investigate the significance of both classic enrichment parameters and advanced descriptors for the performance of 3D virtual screening methods. Furthermore, we review the significance and suitability of RMSD as a measure for the accuracy of protein-ligand docking algorithms and of conformational space sub sampling algorithms.

Virtual screening for the discovery of bioactive natural products

In this survey the impact of the virtual screening concept is discussed in the field of drug discovery from nature. Confronted by a steadily increasing number of secondary metabolites and a growing number of molecular targets relevant in the therapy of human disorders, the huge amount of information needs to be handled. Virtual screening filtering experiments already showed great promise for dealing with large libraries of potential bioactive molecules. It can be utilized for browsing databases for molecules fitting either an established pharmacophore model or a three dimensional (3D) structure of a macromolecular target. However, for the discovery of natural lead candidates the application of this in silico tool has so far almost been neglected. There are several reasons for that. One concerns the scarce availability of natural product (NP) 3D databases in contrast to synthetic libraries; another reason is the problematic compatibility of NPs with modern robotized high throughput screening (HTS) technologies. Further arguments deal with the incalculable availability of pure natural compounds and their often too complex chemistry. Thus research in this field is time-consuming, highly complex, expensive and ineffective. Nevertheless, naturally derived compounds are among the most favorable source of drug candidates. A more rational and economic search for new lead structures from nature must therefore be a priority in order to overcome these problems. Here we demonstrate some basic principles, requirements and limitations of virtual screening strategies and support their applicability in NP research with already performed studies. A sensible exploitation of the molecular diversity of secondary metabolites however asks for virtual screening concepts that are interfaced with well-established strategies from classical pharmacognosy that are used in an effort to maximize their efficacy in drug discovery. Such integrated virtual screening workflows are outlined here and shall help to motivate NP researchers to dare a step towards this powerful in silico tool.

CAESAR: a new conformer generation algorithm based on recursive buildup and local rotational symmetry consideration

A highly efficient conformer search algorithm based on a divide-and-conquer and recursive conformer build-up approach is presented in this paper. This approach is combined with consideration of local rotational symmetry so that conformer duplicates due to topological symmetry in the systematic search can be efficiently eliminated. This new algorithm, termed CAESAR (Conformer Algorithm based on Energy Screening and Recursive Buildup), has been implemented in Discovery Studio 1.7 as part of the Catalyst Component Collection. CAESAR has been validated by comparing the conformer models generated by the new method and Catalyst/FAST. CAESAR is consistently 5-20 times faster than Catalyst/FAST for all data sets investigated. The speedup is even more dramatic for molecules with high topological symmetry or for molecules that require a large number of conformers to be sampled. The quality of the conformer models generated by CAESAR has been validated by assessing the ability to reproduce the receptor-bound X-ray conformation of ligands extracted for the Protein Data Bank (PDB) and assessing the ability to adequately cover the pharmacophore space. It is shown that CAESAR is able to reproduce the receptor-bound conformation slightly better than the Catalyst/FAST method for a data set of 918 ligands retrieved from the PDB. In addition, it is shown that CEASAR covers the pharmacophore space as well or better than Catalyst/FAST.

On the Effects of Permuted Input on Conformational Sampling of Drug-like Molecules: an Evaluation of Stochastic Proximity Embedding

Conformational sampling is a problem of central importance in computer-aided drug design. A good conformational search method must not exhibit any intrinsic bias, and must provide confidence that important regions of conformational space are not missed during the search. A recent study by Carta et al. showed that this is not always the case, and that several popular conformational search methods, such as Omega, are very sensitive to the relative ordering of atoms and bonds in the connection table. Here, we examine the performance of a newer method known as stochastic proximity embedding, or SPE, using five diverse bioactive ligands extracted from the PDB. Our results confirm that the conformational ensembles produced by SPE using different permuted inputs are statistically indistinguishable, and well within the range of variability that would be expected from the stochastic nature of the method itself. This, along with the results of a more comprehensive comparative study (Agrafiotis et al., J. Chem. Info. Model, 2007, in press), provides further evidence that SPE is one of the most robust and competitive conformational search methods described to date.

Assessment of conformational ensemble sizes necessary for specific resolutions of coverage of conformational space

The size of conformational ensembles required for regular coverage of the conformational space of druglike molecules was examined. Using the conformer generation program Omega, the number of regularly distributed conformers (NRC) of flexible compounds was determined as a function of the root-mean-square deviation (RMSD) resolution of coverage. A regression equation was developed predicting the NRC of a molecule as a function of RMSD. The model yielded R(2) of 0.91 for both training and test sets, which consisted of 3414 and 3352 compounds, respectively. Utilizing 14 504 ligands from the Protein Data Bank with experimentally determined 3-D conformations, the regression equation was applied to the estimation of the NRC and the success rate of reproduction of experimental conformations from a theoretical conformation ensemble as a function of RMSD and flexibility was explored.

DNA minor groove pharmacophores describing sequence specific properties

The more that is known about human and other genome sequences and the correlation between gene expression and the course of a disease, the more evident it seems to be that DNA is chosen as a drug target instead of proteins which are built with the information encoded by DNA. According to this approach, small minor groove binding molecules have been designed to bind the DNA sequence specifically and thereby downregulate genes. Because of their lack of druglikeness, we plan to use them as templates for forthcoming virtual screening experiments to discover molecules with the same bioactivity and a different scaffold. In this proof of principle study, carried out with the software tool Catalyst, we present a model work for description of a ligand-DNA complex with the aid of pharmacophore modeling methods. The successful reproduction of sequence specificity of a polyamidic minor groove binding ligand is the precondition for later model application to virtual screening.

Recent Advances in Chemoinformatics

Chemoinformatics is a large scientific discipline that deals with the storage, organization, management, retrieval, analysis, dissemination, visualization, and use of chemical information. Chemoinformatics techniques are used extensively in drug discovery and development. Although many consider it a mature field, the advent of high-throughput experimental techniques and the need to analyze very large data sets have brought new life and challenges to it. Here, we review a selection of papers published in 2006 that caught our attention with regard to the novelty of the methodology that was presented. The field is seeing significant growth, which will be further catalyzed by the widespread availability of public databases to support the development and validation of new approaches.

Conformational coverage by a genetic algorithm: saturation of conformational space

The molecular modeling is traditionally based on analysis of minimum energy conformers. Such simplifying assumptions could doom to failure the modeling studies given the significant variation of the geometric and electronic characteristics across the multitude of energetically reasonable conformers representing the molecules. Moreover, it has been found that the lowest energy conformers of chemicals are not necessarily the active ones with respect to various endpoints. Hence, the selection of active conformers appears to be as important as the selection of molecular descriptors in the modeling process. In this respect, we have developed effective tools for conformational analysis based on a genetic algorithm (GA), published in J. Chem. Inf. Comput. Sci. (1994, 34, 234; 1999, 39 (6), 997) and J. Chem. Inf. Model. (2005, 45 (2), 283). This paper presents a further improvement of the evolutionary algorithm for conformer generation minimizing the sensitivity of conformer distributions from the effect of smoothing parameter and improving the reproducibility of conformer distributions given the nondeterministic character of the genetic algorithm (GA). The ultimate goal of the saturation is to represent the conformational space of chemicals with an optimal number of conformers providing a stable conformational distribution which cannot be further perturbed by the addition of new conformers. The generation of stable conformational distributions of chemicals by a limited number of conformers will improve the adequacy of the subsequent molecular modeling analysis. The impact of the saturation procedure on conformer distributions in a specific structural space is illustrated by selected examples. The effect of the procedure on similarity assessment between chemicals is discussed.

Conformational sampling of bioactive molecules: a comparative study

The necessity to generate conformations that sample the entire conformational space accessible to a given molecule is ubiquitous in the field of computer-aided drug design. Protein-ligand docking, 3D database searching, and 3D QSAR are three commonly used techniques that depend critically upon the quality and diversity of the generated conformers. Although there are a wide range of conformational search algorithms available, the extent to which they sample conformational space is often unclear. To address this question, we conducted a robust comparison of the search algorithms implemented in several widely used molecular modeling packages, including Catalyst, Macromodel, Omega, MOE, and Rubicon as well as our own method, stochastic proximity embedding (SPE). We found that SPE used in conjunction with conformational boosting, a heuristic for biasing conformational search toward more extended or compact geometries, along with Catalyst, are significantly more effective in sampling the full range of conformational space compared to the other methods, which show distinct preferences for either more extended or more compact geometries.

Parallel screening: a novel concept in pharmacophore modeling and virtual screening

Parallel screening comprises a novel in silico method to predict the potential biological activities of a compound by screening it with a multitude of pharmacophore models. Our aim is to provide a fast, large-scale system that allows for virtual activity profiling. In this proof of principle study, carried out with the software tools LigandScout and Catalyst, we present a model work for the application of parallel pharmacophore-based virtual screening on a set of 50 structure-based pharmacophore models built for various viral targets and 100 antiviral compounds. The latter were screened against all pharmacophore models in order to determine if their biological targets could be correctly predicted via an enrichment of corresponding pharmacophores matching these ligands. The results demonstrate that the desired enrichment, that is, successful virtual activity profiling, was achieved for approximately 90% of all input molecules. We discuss descriptors for output validation, as well as various aspects influencing the analysis of the obtained activity profiles, and the effect of the utilized search modus for screening.

Efficient overlay of small organic molecules using 3D pharmacophores

Aligning and overlaying two or more bio-active molecules is one of the key tasks in computational drug discovery and bio-activity prediction. Especially chemical-functional molecule characteristics from the view point of a macromolecular target represented as a 3D pharmacophore are the most interesting similarity measure when describing and analyzing macromolecule-ligand interaction. In this study, a novel approach for aligning rigid three-dimensional molecules according to their chemical-functional pharmacophoric features is presented and compared to the overlay of experimentally determined poses in a comparable macromolecule coordinate frame. The presented approach identifies optimal chemical feature pairs using distance and density characteristics obtained by correlating pharmacophoric geometries and thus proves to be faster than existing combinatorial alignment methods and creates more reasonable alignments than pure atom-based methods. Examples will be provided to demonstrate the feasibility, speed and intuitiveness of this method.

High-throughput structure-based pharmacophore modelling as a basis for successful parallel virtual screening

In order to assess bioactivity profiles for small organic molecules we propose to use parallel pharmacophore-based virtual screening. Our aim is to provide a fast, reliable and scalable system that allows for rapid in silico activity profile prediction of virtual molecules. In this proof of principle study, carried out with the new structure-based pharmacophore modelling tool LigandScout and the high-performance database mining platform Catalyst, we present a model work for the application of parallel pharmacophore-based virtual screening on a set of 50 structure-based pharmacophore models built for various viral targets and 100 antiviral compounds. The latter were screened against all pharmacophore models in order to determine if their known biological targets could be correctly predicted via an enrichment of corresponding pharmaco-phores matching these ligands. The results demonstrate that the desired enrichment, i.e. a successful activity profiling, was achieved for approximately 90% of all input molecules. Additionally, we discuss descriptors for output validation, as well as various aspects influencing the analysis of the obtained activity profiles, and the effect of the searching mode utilized for screening. The results of the study presented here clearly indicate that pharmacophore-based parallel screening comprises a reliable in silico method to predict the potential biological activities of a compound or a compound library by screening it against a series of pharmacophore queries.

Comparative performance assessment of the conformational model generators omega and catalyst: a large-scale survey on the retrieval of protein-bound ligand conformations

In continuation of our studies to evaluate the ability of various conformer generators to produce bioactive conformations, we present the extension of our work on the analysis of Catalyst's conformational subsampling algorithm in a comparative evaluation with OpenEye's currently updated tool Omega 2.0. Our study is based on an enhanced test set of 778 drug molecules and pharmacologically relevant compounds extracted from the Protein Data Bank (PDB). We elaborated protocols for two common conformer generation use cases and applied them to both programs: (i) high-throughput settings for processing large databases and (ii) high-quality settings for binding site exploration or lead structure refinement. While Catalyst is faster in the first case, Omega 2.0 better reproduces the bound ligand conformations from the PDB in less time for the latter case.

Using pharmacophore models to gain insight into structural binding and virtual screening: an application study with CDK2 and human DHFR

This study provides results from two case studies involving the application of the HypoGenRefine algorithm within Catalyst for the automated generation of excluded volume from ligand information alone. A limitation of pharmacophore feature hypothesis alone is that activity prediction is based purely on the presence and arrangement of pharmacophoric features; steric effects remained unaccounted. Recently reported studies have illustrated the usefulness of combining excluded volumes to the pharmacophore models. In general, these excluded volumes attempt to penalize molecules occupying steric regions that are not occupied by active molecules. The HypoGenRefine algorithm in Catalyst accounts for steric effects on activity, based on the targeted addition of excluded volume features to the pharmacophores. The automated inclusion of excluded volumes to pharmacophore models has been applied to two systems: CDK2 and human DHFR. These studies are used as examples to illustrate how ligands could bind in the protein active site with respect to allowed and disallowed binding regions. Additionally, automated refinement of the pharmacophore with these excluded volume features provides a more selective model to reduce false positives and a better enrichment rate in virtual screening.

Influence of the Conditions in Pharmacophore Generation, Scoring, and 3D Database Search for Chemical Feature-Based Pharmacophore Models: One Application Study of ETA- and ETB-Selective Antagonists

Using the commercial pharmacophore modeling suite Catalyst, we have studied the influence of the compare.scaledMultiBlobFeatureErrors . Catalyst parameter. The influence of this parameter has been studied in pharmacophore generation, hypothesis scoring, and database searching. This parameter, introduced in Catalyst 4.7, changed its default value in Catalyst 4.8, and it strongly influences the statistical quality of pharmacophore generation, scoring of the hypotheses, and database searching. Two different pharmacophore models have been constructed for the ETA and ETB receptor antagonists. Both models contain one positive ionizable, one negative ionizable, one hydrogen-bond acceptor, one hydrophobic aromatic, and one hydrophobic aliphatic feature. The models have been compared, and some differences in the position of the hydrogen-bond acceptor in the putative binding pocket have been highlighted.