A fast new approach to pharmacophore mapping and its application to dopaminergic and benzodiazepine agonists

Summary In the absence of a 3D structure of the target biomolecule, to propose the 3D requirements for a small molecule to exhibit a particular bioactivity, one must supply both a bioactive conformation and a superposition rule for every active compound. Our strategy identifies both simultaneously. We first generate and optimize all low-energy conformations by any suitable method. For each conformation we then use ALAD-DIN to calculate the location of points to be considered as part of the superposition. These points include atoms in the molecule and projections from the molecule to hydrogen-bond donors and acceptors or charged groups in the binding site. These positions and the relative energy of each conformation are the input to our new program DISCO. It uses a clique-detection method to find superpositions that contain a least one conformation of each molecule and user-specified numbers of point types and chirality. DISCO is fast; for example, it takes about 1 min CPU to propose pharmacophores from 21 conformations of seven molecules. We typically run DISCO several times to compare alternative pharmacophore maps. For D₂ dopamine agonists DISCO shows that the newer 2-aminothiazoles fit the traditional pharmacophore. Using site points correctly identifies the bioactive enantiomers of indoles to compare with catechols whereas using only ligand points leads to selecting the inactive enantiomer for the pharmacophore map. In addition, DISCO reproduces pharmacophore maps of benzodiazepines in the literature and proposes subtle improvements. Our experience suggests that clique-detection methods will find many applications in computational chemistry and computer-assisted molecular design.     

A marriage made in torsional space: using GALAHAD models to drive pharmacophore multiplet searches

Pharmacophore multiplets are useful tools for 3D database searching, with the queries used ordinarily being derived from ensembles of random conformations of active ligands. It seems reasonable to expect that their usefulness can be augmented by instead using queries derived from single ligand conformations obtained from aligned ligands. Comparisons of pharmacophore multiplet searching using random conformations with multiplet searching using single conformations derived from GALAHAD (a genetic algorithm with linear assignment for hypermolecular alignment of datasets) models do indeed show that, while query hypotheses based on random conformations are quite effective, hypotheses based on aligned conformations do a better job of discriminating between active and inactive compounds. In particular, the hypothesis created from a neuraminidase inhibitor model was more similar to half of 18 known actives than all but 0.2% of the compounds in a structurally diverse subset of the World Drug Index. Similarly, a model developed from five angiotensin II antagonists yielded hypotheses that placed 65 known antagonists within the top 0.1–1% of decoy databases. The differences in discriminating power ranged from 2 to 20-fold, depending on the protein target and the type of pharmacophore multiplet used.     

Modeling ligand-receptor interaction for some MHC class II HLA-DR4 peptide mimetic inhibitors using several molecular docking and 3D QSAR techniques

The ligand-receptor interaction between some peptidomimetic inhibitors and a class II MHC peptide presenting molecule, the HLA-DR4 receptor, was modeled using some three-dimensional (3D) quantitative structure-activity relationship (QSAR) methods such as the Comparative Molecular Field Analysis (CoMFA), Comparative Molecular Similarity Indices Analysis (CoMSIA), and a pharmacophore building method, the Catalyst program. The structures of these peptidomimetic inhibitors were generated theoretically, and the conformations used in the 3D QSAR studies were defined by docking them into the known structure of HLA-DR4 receptor through the GOLD, GLIDE Rigidly, GLIDE Flexible, and Xscore programs. Some of the parameters used in these docking programs were selected by docking an X-ray ligand into the receptor and comparing the root-means-square difference (RMSD) computed between the coordinates of the X-ray and docked structure. However, the goodness of a docking result for docking a series of peptidomimetic inhibitors into the HLA-DR4 receptor was judged by comparing the Spearman's rank correlation coefficient computed between each docking result and the activity data taken from the literature. The best CoMFA and CoMSIA models were constructed using the aligned structures of the best docking result. The CoMSIA was conducted in a stepwise manner to identify some important molecular features that were further employed in a pharmacophore building process by the Catalyst program. It was found that most inhibitors of the training set were accurately predicted by the best pharmacophore model, the Hypo1 hypothesis constructed. The deviation or conflict found between the actual and predicted activities of some inhibitors of both the training and the test sets were also investigated by mapping the Hypo1 hypothesis onto the corresponding structures of the inhibitors.     

Greedy 3-Point Search (G3PS)—A Novel Algorithm for Pharmacophore Alignment

Chemical features of small molecules can be abstracted to 3D pharmacophore models, which are easy to generate, interpret, and adapt by medicinal chemists. Three-dimensional pharmacophores can be used to efficiently match and align molecules according to their chemical feature pattern, which facilitates the virtual screening of even large compound databases. Existing alignment methods, used in computational drug discovery and bio-activity prediction, are often not suitable for finding matches between pharmacophores accurately as they purely aim to minimize RMSD or maximize volume overlap, when the actual goal is to match as many features as possible within the positional tolerances of the pharmacophore features. As a consequence, the obtained alignment results are often suboptimal in terms of the number of geometrically matched feature pairs, which increases the false-negative rate, thus negatively affecting the outcome of virtual screening experiments. We addressed this issue by introducing a new alignment algorithm, Greedy 3-Point Search (G3PS), which aims at finding optimal alignments by using a matching-feature-pair maximizing search strategy while at the same time being faster than competing methods.     

Estimation of active conformations of drugs by a new molecular superposing procedure

We have developed a new program, SUPERPOSE, to superpose two molecules based on the physicochemical properties of functional atoms within individual molecules. SUPERPOSE treats a pseudo-molecule consisting of functional atoms instead of a real molecule. Four types of physicochemical properties – hydrophobicity, presence of a hydrogen-bonding donor, presence of a hydrogen-bonding acceptor and presence of a hydrogen-bonding donor/acceptor – were supposed and a score was given to each overlap. When functional atoms with the same physicochemical properties were overlapped, points were added to the score, and when the functional atoms with different physicochemical properties were overlapped, points were subtracted. We applied SUPERPOSE to 12 pairs of 24 enzyme inhibitors and found that the best scored overlay for each inhibitor pair could successfully reproduce the superposition obtained from X-ray crystallography. Next, we applied SUPERPOSE to estimate the active conformations of the thrombin inhibitors MQPA, 4-TAPAP and NAPAP. Superpositions of conformers sampled by the high-temperature molecular dynamics calculation with respect to the three inhibitors were performed, and 13 sets of conformers having the best common overlay to the three inhibitors were selected. One among 13 sets was consistent with the superposition of the active conformations derived from the X-ray crystallography of the thrombin–inhibitor complexes.     

Modeling flexible pharmacophores with distance geometry, scoring, and bound stretching

The study of pharmacophores, i.e., of common features between different ligands, is important for the quantitative identification of "compatible" enzymes and binding species. A pharmacophore-based technique is developed that combines multiple conformations with a distance geometry method to create flexible pharmacophore representations. It uses a set of low-energy conformations combined with a new process we call bound stretching to create sets of distance bounds, which contain all or most of the low-energy conformations. The bounds can be obtained using the exact distances between pairs of atoms from the different low-energy conformations. To avoid missing conformations, we can take advantage of the triangle distance inequality between sets of three points to logically expand a set of upper and lower distance bounds (bound stretching). The flexible pharmacophore can be found using a 3-D maximal common subgraph method, which uses the overlap of distance bounds to determine the overlapping structure. A scoring routine is implemented to select the substructures with the largest overlap because there will typically be many overlaps with the maximum number of overlapping bounds. A case study is presented in which 3-D flexible pharmacophores are generated and used to eliminate potential binding species identified by a 2-D pharmacophore method. A second case study creates flexible pharmacophores from a set of thrombin ligands. These are used to compare the new method with existing pharmacophore identification software.     

A ligand-based molecular modeling study on some matrix metalloproteinase-1 inhibitors using several 3D QSAR techniques

Some three-dimensional quantitative structure-activity relationship (3D-QSAR) models have been constructed using the comparative molecular field analysis (CoMFA) and comparative molecular similarity indices (CoMSIA) for a series of 84 proline-based plus 12 structurally more diversified nonproline matrix metalloproteinase inhibitors. The structures of these inhibitors were built from a structure template extracted from the crystal structure of stromelysin. The structures built were divided into the training and test sets for both the CoMFA and CoMSIA analyses for each being composed of 60 and 24 inhibitors, respectively. The structures in the training set were aligned using some alignment rules derived from the analysis of the Ligplot program on a recent crystal structure of ligand-collagenase-1 complex. Some stepwise CoMSIA's were performed on the aligned training set on which the best CoMFA result was obtained. The best CoMSIA model was identified from the stepwise results, and the corresponding pharmacophore features were used for the construction of a pharmacophore hypothesis by the Catalyst 4.9 program. The training set was extended to include 11 structurally more diversified and nonproline inhibitors. To construct a pharmacophore hypothesis, the conformation of 60 structurally aligned proline-based inhibitors was fixed, while that of the 11 structurally more diversified nonproline inhibitors was allowed to vary during the hypothesis construction process. It was found that the predicted activities by the top hypothesis constructed for both the training and test sets were as good in statistics as those predicted by the best CoMSIA model from which the hypothesis was derived. The top hypothesis was mapped onto the structures of several highly active inhibitors selected from both the training and test sets. The goodness of mapping on each inhibitor was found to be correlated well with the activity of each inhibitor.     

Flexible matching of test ligands to a 3D pharmacophore using a molecular superposition force field: Comparison of predicted and experimental conformations of inhibitors of three enzymes

Summary A computer procedure TFIT, which uses a molecular superposition force field to flexibly match test compounds to a 3D pharmacophore, was evaluated to find out whether it could reliably predict the bioactive conformations of flexible ligands. The program superposition force field optimizes the overlap of those atoms of the test ligand and template that are of similar chemical type, by applying an attractive force between atoms of the test ligand and template which are close together and of similar type (hydrogen bonding, charge, hydrophobicity). A procedure involving Monte Carlo torsion perturbations, followed by torsional energy minimization, is used to find conformations of the test ligand which cominimize the internal energy of the ligand and the superposition energy of ligand and template. The procedure was tested by applying it to a series of flexible ligands for which the bioactive conformation was known experimentally. The 15 molecules tested were inhibitors of thermolysin, HIV-1 protease or endothiapepsin for which X-ray structures of the bioactive conformation were available. For each enzyme, one of the molecules served as a template and the others, after being conformationally randomized, were fitted. The fitted conformation was then compared to the known binding geometry. The matching procedure was successful in predicting the bioactive conformations of many of the structures tested. Significant deviation from experimental results was found only for parts of molecules where it was readily apparent that the template did not contain sufficient information to accurately determine the bioactive conformation.     

Geometric accuracy of three-dimensional molecular overlays

This study examines the dependence of molecular alignment accuracy on a variety of factors including the choice of molecular template, alignment method, conformational flexibility, and type of protein target. We used eight test systems for which X-ray data on 145 ligand-protein complexes were available. The use of X-ray structures allowed an unambiguous assignment of bioactive overlays for each compound set. The alignment accuracy depended on multiple factors and ranged from 6% for flexible overlays to 73% for X-ray rigid overlays, when the conformation of the template ligand came from X-ray structures. The dependence of the overlay accuracy on the choice of templates and molecules to be aligned was found to be the most significant factor in six and seven of the eight ligand-protein complex data sets, respectively. While finding little preference for the overlay method, we observed that the introduction of molecule flexibility resulted in a decrease of overlay accuracy in 50% of the cases. We derived rules to maximize the accuracy of alignment, leading to a more than 2-fold improvement in accuracy (from 19% to 48%). The rules also allowed the identification of compounds with a low (<5%) chance to be correctly aligned. Last, the accuracy of the alignment derived without any utilization of X-ray conformers varied from <1% for the human immunodeficiency virus data set to 53% for the trypsin data set. We found that the accuracy was directly proportional to the product of the overlay accuracy from the templates in their bioactive conformations and the chance of obtaining the correct bioactive conformation of the templates. This study generates a much needed benchmark for the expectations of molecular alignment accuracy and shows appropriate usages and best practices to maximize hypothesis generation success.     

Graph-based molecular alignment (GMA)

We describe a combined 2D/3D approach for the superposition of flexible chemical structures, which is based on recent progress in the efficient identification of common subgraphs and a gradient-based torsion space optimization algorithm. The simplicity of the approach is reflected in its generality and computational efficiency: the suggested approach neither requires precalculated statistics on the conformations of the molecules nor does it make simplifying assumptions on the topology of the molecules being compared. Furthermore, graph-based molecular alignment produces alignments that are consistent with the chemistry of the molecules as well as their general structure, as it depends on both the local connectivities between atoms and the overall topology of the molecules. We validate this approach on benchmark sets taken from the literature and show that it leads to good results compared to computationally and algorithmically more involved methods. The results suggest that, for most practical purposes, graph-based molecular alignment is a viable alternative to molecular field alignment with respect to structural superposition and leads to structures of comparable quality in a fraction of the time.     

A hybrid approach for addressing ring flexibility in 3D database searching

A hybrid approach for flexible 3D database searching is presented that addresses the problemof ring flexibility. It combines the explicit storage of up to 25 multiple conformations ofrings, with up to eight atoms, generated by the 3D structure generator CORINA with thepower of a torsional fitting technique implemented in the 3D database system UNITY. Acomparison with the original UNITY approach, using a database with about 130,000 entriesand five different pharmacophore queries, was performed. The hybrid approach scored, on anaverage, 10–20% more hits than the reference run. Moreover, specific problems withunrealistic hit geometries produced by the original approach can be excluded. In addition, theinfluence of the maximum number of ring conformations per molecule was investigated. Anoptimal number of 10 conformations per molecule is recommended.     

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.     

Unsupervised 3D ring template searching as an ideas generator for scaffold hopping: use of the LAMDA, RigFit, and field-based similarity search (FBSS) methods

Crystal structures taken from the Cambridge Structural Database were used to build a ring scaffold database containing 19 050 3D structures, with each such scaffold then being used to generate a centroid connecting path (CCP) representation. The CCP is a novel object that connects ring centroids, ring linker atoms, and other important points on the connection path between ring centroids. Unsupervised searching in the scaffold and CCP data sets was carried out using the atom-based LAMDA and RigFit search methods and the field-based similarity search method. The performance of these methods was tested with three different ring scaffold queries. These searches demonstrated that unsupervised 3D scaffold searching methods can find not only the types of ring systems that might be retrieved in carefully defined pharmacophore searches (supervised approach) but also additional, structurally diverse ring systems that could form the starting point for lead discovery programs or other scaffold-hopping applications. Not only are the methods effective but some are sufficiently rapid to permit scaffold searching in large chemical databases on a routine basis.     

虚拟活性化合物的自动生成

虚拟活性化合物的自动生成是从药效团和预先设定的结构碎片出发,通过碎片选择、碎片组装和柔性构象搜索来获得虚拟活性结构。经过对HIV-1蛋白酶抑制剂药效团进行虚拟活性化合物生成,得到了16个虚拟活性化合物,通过构象分析发现生成的化合物满足药效团的限制条件。说明这一方法能够有效的生成虚拟活性结构,与药效团检索结果对比发现生成的虚拟活性结构新颖易于合成。     

A randomized kinematics‐based approach to pharmacophore‐constrained conformational search and database screening

Computational tools have greatly expedited the pharmaceutical drug design process in recent years. One common task in this process is the search of a large library for small molecules that can achieve both a low‐energy conformation and a prescribed pharmacophore. The pharmacophore expresses constraints on the 3D structure of the molecule by specifying relative atom positions that should be maintained to increase the likelihood that the molecule will bind with the receptor site. This article presents a pharmacophore‐based database screening system that has been designed, implemented, and tested on a molecular database. The primary focus of this article is on a simple, randomized conformational search technique that attempts to simultaneously reduce energy and maintain pharmacophore constraints. This enables the identification of molecules in a database that are likely to dock with a given protein, which can serve as a powerful aid in the search for better drug candidates. © 2000 John Wiley & Sons, Inc. J Comput Chem 21: 731–747, 2000