GNE-064: A Potent, Selective, and Orally Bioavailable Chemical Probe for the Bromodomains of SMARCA2 and SMARCA4 and the Fifth Bromodomain of PBRM1

Bromodomains are acetyllysine recognition domains present in a variety of human proteins. Bromodomains also bind small molecules that compete with acetyllysine, and therefore bromodomains have been targets for drug discovery efforts. Highly potent and selective ligands with good cellular permeability have been proposed as chemical probes for use in exploring the functions of many of the bromodomain proteins. We report here the discovery of a class of such inhibitors targeting the family VIII bromodomains of SMARCA2 (BRM) and SMARCA4 (BRG1), and PBRM1 (polybromo-1) bromodomain 5. We propose one example from this series, GNE-064, as a chemical probe for the bromodomains SMARCA2, SMARCA4, and PBRM1(5) with the potential for in vivo use.

Design and Synthesis of Styrenylcyclopropylamine LSD1 Inhibitors

Leveraging the catalytic machinery of LSD1 (KDM1A), a series of covalent styrenylcyclopropane LSD1 inhibitors were identified. These inhibitors represent a new class of mechanism-based inhibitors that target and covalently label the FAD cofactor of LSD1. The series was rapidly progressed to potent biochemical and cellular LSD1 inhibitors with good physical properties. This effort resulted in the identification of 34, a highly potent (<4 nM biochemical, 2 nM cell, and 1 nM GI50), and selective LSD1 inhibitor. In-depth kinetic profiling of 34 confirmed its covalent mechanism of action, validated the styrenylcyclopropane as an FAD-directed warhead, and demonstrated that the potency of this inhibitor is driven by improved non-covalent binding (K I). 34 demonstrated robust cell-killing activity in a panel of AML cell lines and robust antitumor activity in a Kasumi-1 xenograft model of AML when dosed orally at 1.5 mg/kg once daily.

P3-P4 ureas and reverse carbamates as potent HCV NS3 protease inhibitors: Effective transposition of the P4 hydrogen bond donor

A series of tripeptidic acylsulfonamide inhibitors of HCV NS3 protease were prepared that explored structure-activity relationships (SARs) at the P4 position, and their in vitro and in vivo properties were evaluated. Enhanced potency was observed in a series of P4 ureas; however, the PK profiles of these analogues were less than optimal. In an effort to overcome the PK shortcomings, modifications to the P3-P4 junction were made. This included a strategy in which one of the two urea N-H groups was either N-methylated or replaced with an oxygen atom. The former approach provided a series of regioisomeric N-methylated ureas while the latter gave rise to P4 reverse carbamates, both of which retained potent NS3 inhibitory properties while relying upon an alternative H-bond donor topology. Details of the SARs and PK profiles of these analogues are provided.

Bicyclic Helical Peptides as Dual Inhibitors Selective for Bcl2A1 and Mcl-1 Proteins

A 26-residue peptide BimBH3 binds indiscriminately to multiple oncogenic Bcl2 proteins that regulate apoptosis of cancer cells. Specific inhibition of the BimBH3-Bcl2A1 protein-protein interaction was obtained in vitro and in cancer cells by shortening the peptide to 14 residues, inserting two cyclization constraints to stabilize a water-stable α-helix, and incorporating an N-terminal acrylamide electrophile for selective covalent bonding to Bcl2A1. Mass spectrometry of trypsin-digested bands on electrophoresis gels established covalent bonding of an electrophilic helix to just one of the three cysteines in Bcl2A1, the one (Cys55) at the BimBH3-Bcl2A1 protein-protein interaction interface. Optimizing the helix-inducing constraints and the sequence subsequently enabled electrophile removal without loss of inhibitor potency. The bicyclic helical peptides were potent, cell permeable, plasma-stable, dual inhibitors of Bcl2A1 and Mcl-1 with high selectivity over other Bcl2 proteins. One bicyclic peptide was shown to inhibit the interaction between a pro-apoptotic protein (Bim) and either endogenous Bcl2A1 or Mcl-1, to induce apoptosis of SKMel28 human melanoma cells, and to sensitize them for enhanced cell death by the anticancer drug etoposide. These approaches look promising for chemically silencing intracellular proteins.

Potent Inhibitors of Hepatitis C Virus NS3 Protease: Employment of a Difluoromethyl Group as a Hydrogen-Bond Donor

The design and synthesis of potent, tripeptidic acylsulfonamide inhibitors of HCV NS3 protease that contain a difluoromethyl cyclopropyl amino acid at P1 are described. A cocrystal structure of 18 with a NS3/4A protease complex suggests the presence of a H-bond between the polarized C-H of the CHF₂ moiety and the backbone carbonyl of Leu135 of the enzyme. Structure-activity relationship studies indicate that this H-bond enhances enzyme inhibitory potency by 13- and 17-fold compared to the CH₃ and CF₃ analogues, respectively, providing insight into the deployment of this unique amino acid.

Electrophilic Helical Peptides That Bond Covalently, Irreversibly, and Selectively in a Protein-Protein Interaction Site

Protein-protein interactions mediate most physiological and disease processes. Helix-constrained peptides potently mimic or inhibit these interactions by making multiple contacts over large surface areas. However, despite high affinities, they typically have short lifetimes bound to the protein. Here we insert both a helix-inducing constraint and an adjacent electrophile into the native peptide ligand BIM to target the oncogenic protein Bcl2A1. The modified BIM peptide bonds covalently and irreversibly to one cysteine within the helix-binding groove of Bcl2A1, but not to two other exposed cysteines on its surface, and shows no covalent bonding to other Bcl2 proteins. It also penetrates cell membranes and bonds covalently to Bcl2A1 inside cells. This innovative approach to increasing receptor residence time of helical peptides demonstrates the potential to selectively silence a PPI inside cells, with selectivity over other nucleophilic sites on proteins.

Identification of (R)-N-((4-Methoxy-6-methyl-2-oxo-1,2-dihydropyridin-3-yl)methyl)-2-methyl-1-(1-(1-(2,2,2-trifluoroethyl)piperidin-4-yl)ethyl)-1H-indole-3-carboxamide (CPI-1205), a Potent and Selective Inhibitor of Histone Methyltransferase EZH2, Suitable for Phase I Clinical Trials for B-Cell Lymphomas

Polycomb repressive complex 2 (PRC2) has been shown to play a major role in transcriptional silencing in part by installing methylation marks on lysine 27 of histone 3. Dysregulation of PRC2 function correlates with certain malignancies and poor prognosis. EZH2 is the catalytic engine of the PRC2 complex and thus represents a key candidate oncology target for pharmacological intervention. Here we report the optimization of our indole-based EZH2 inhibitor series that led to the identification of CPI-1205, a highly potent (biochemical IC₅₀ = 0.002 μM, cellular EC₅₀ = 0.032 μM) and selective inhibitor of EZH2. This compound demonstrates robust antitumor effects in a Karpas-422 xenograft model when dosed at 160 mg/kg BID and is currently in Phase I clinical trials. Additionally, we disclose the co-crystal structure of our inhibitor series bound to the human PRC2 complex.

Diving into the Water: Inducible Binding Conformations for BRD4, TAF1(2), BRD9, and CECR2 Bromodomains

The biological role played by non-BET bromodomains remains poorly understood, and it is therefore imperative to identify potent and highly selective inhibitors to effectively explore the biology of individual bromodomain proteins. A ligand-efficient nonselective bromodomain inhibitor was identified from a 6-methyl pyrrolopyridone fragment. Small hydrophobic substituents replacing the N-methyl group were designed directing toward the conserved bromodomain water pocket, and two distinct binding conformations were then observed. The substituents either directly displaced and rearranged the conserved solvent network, as in BRD4(1) and TAF1(2), or induced a narrow hydrophobic channel adjacent to the lipophilic shelf, as in BRD9 and CECR2. The preference of distinct substituents for individual bromodomains provided selectivity handles useful for future lead optimization efforts for selective BRD9, CECR2, and TAF1(2) inhibitors.

Discovery, design, and synthesis of indole-based EZH2 inhibitors

The discovery and optimization of a series of small molecule EZH2 inhibitors is described. Starting from dimethylpyridone HTS hit (2), a series of indole-based EZH2 inhibitors were identified. Biochemical potency and microsomal stability were optimized during these studies and afforded compound 22. This compound demonstrates nanomolar levels of biochemical potency (IC50=0.002 μM), cellular potency (EC50=0.080 μM), and afforded tumor regression when dosed (200 mpk SC BID) in an EZH2 dependent tumor xenograft model.

The discovery of asunaprevir (BMS-650032), an orally efficacious NS3 protease inhibitor for the treatment of hepatitis C virus infection

The discovery of asunaprevir (BMS-650032, 24) is described. This tripeptidic acylsulfonamide inhibitor of the NS3/4A enzyme is currently in phase III clinical trials for the treatment of hepatitis C virus infection. The discovery of 24 was enabled by employing an isolated rabbit heart model to screen for the cardiovascular (CV) liabilities (changes to HR and SNRT) that were responsible for the discontinuation of an earlier lead from this chemical series, BMS-605339 (1), from clinical trials. The structure-activity relationships (SARs) developed with respect to CV effects established that small structural changes to the P2* subsite of the molecule had a significant impact on the CV profile of a given compound. The antiviral activity, preclincial PK profile, and toxicology studies in rat and dog supported clinical development of BMS-650032 (24).

Molecular shape and medicinal chemistry: a perspective

The eight contributions here provide ample evidence that shape as a volume or as a surface is a vibrant and useful concept when applied to drug discovery. It provides a reliable scaffold for "decoration" with chemical intuition (or bias) for virtual screening and lead optimization but also has its unadorned uses, as in library design, ligand fitting, pose prediction, or active site description. Computing power has facilitated this evolution by allowing shape to be handled precisely without the need to reduce down to point descriptors or approximate metrics, and the diversity of resultant applications argues for this being an important step forward. Certainly, it is encouraging that as computation has enabled our intuition, molecular shape has consistently surprised us in its usefulness and adaptability. The first Aurelius question, "What is the essence of a thing?", seems well answered, however, the third, "What do molecules do?", only partly so. Are the topics covered here exhaustive, or is there more to come? To date, there has been little published on the use of the volumetric definition of shape described here as a QSAR variable, for instance, in the prediction or classification of activity, although other shape definitions have been successful applied, for instance, as embodied in the Compass program described above in "Shape from Surfaces". Crystal packing is a phenomenon much desired to be understood. Although powerful models have been applied to the problem, to what degree is this dominated purely by the shape of a molecule? The shape comparison described here is typically of a global nature, and yet some importance must surely be placed on partial shape matching, just as the substructure matching of chemical graphs has proved useful. The approach of using surfaces, as described here, offers some flavor of this, as does the use of metrics that penalize volume mismatch less than the Tanimoto, e.g., Tversky measures. As yet, there is little to go on as to how useful a paradigm this will be because there is less software and fewer concrete results.Finally, the distance between molecular shapes, or between any shapes defined as volumes or surfaces, is a metric property in the mathematical sense of the word. As yet, there has been little, if any, application of this observation. We cannot know what new application to the design and discovery of pharmaceuticals may yet arise from the simple concept of molecular shape, but it is fair to say that the progress so far is impressive.

Novel DOCK clique driven 3D similarity database search tools for molecule shape matching and beyond: Adding flexibility to the search for ligand kin

With readily available CPU power and copious disk storage, it is now possible to undertake rapid comparison of 3D properties derived from explicit ligand overlay experiments. With this in mind, shape software tools originally devised in the 1990s are revisited, modified and applied to the problem of ligand database shape comparison. The utility of Connolly surface data is highlighted using the program MAKESITE, which leverages surface normal data to a create ligand shape cast. This cast is applied directly within DOCK, allowing the program to be used unmodified as a shape searching tool. In addition, DOCK has undergone multiple modifications to create a dedicated ligand shape comparison tool KIN. Scoring has been altered to incorporate the original incarnation of Gaussian function derived shape description based on STO-3G atomic electron density. In addition, a tabu-like search refinement has been added to increase search speed by removing redundant starting orientations produced during clique matching. The ability to use exclusion regions, again based on Gaussian shape overlap, has also been integrated into the scoring function. The use of both DOCK with MAKESITE and KIN in database screening mode is illustrated using a published ligand shape virtual screening template. The advantages of using a clique-driven search paradigm are highlighted, including shape optimization within a pharmacophore constrained framework, and easy incorporation of additional scoring function modifications. The potential for further development of such methods is also discussed.

Ranking poses in structure-based lead discovery and optimization: current trends in scoring function development

The ability to accurately predict the potency of ligand binding to its intended target prior to synthesis is of significant value in the drug-discovery paradigm. The protocols designed to this end follow a two-step process. First, ligands are docked into the active site of interest, and then the resulting interactions at the target are scored. Scoring functions form the key component in this process, providing a quantitative measure of fit quality. There is an abundance of new research in the field of scoring function design, from incorporation of novel descriptors (derived from first principles or empirical analysis) to function redesign based on improved data set handling. This article provides a brief overview on the state-of-the-art developments in the field, with particular reference to their performance in relation to expected outcomes.

Measuring CAMD Technique Performance. 2. How “Druglike” Are Drugs? Implications of Random Test Set Selection Exemplified Using Druglikeness Classification Models

Research into the advancement of computer-aided molecular design (CAMD) has a tendency to focus on the discipline of algorithm development. Such efforts are often wrought to the detriment of the data set selection and analysis used in said algorithm validation. Here we highlight the potential problems this can cause in the context of druglikeness classification. More rigorous efforts are applied to the selection of decoy (nondruglike) molecules from the ACD. Comparisons are made between model performance using the standard technique of random test set creation with test sets derived from explicit ontological separation by drug class. The dangers of viewing druglike space as sufficiently coherent to permit simple classification are highlighted. In addition the issues inherent in applying unfiltered data and random test set selection to (Q)SAR models utilizing large and supposedly heterogeneous databases are discussed.

An Empirical Process for the Design of High-Throughput Screening Deck Filters

A process for objective identification and filtering of undesirable compounds that contribute to high-throughput screening (HTS) deck promiscuity is described. Two methods of mapping hit promiscuity have been developed linking SMARTS-based structural queries with historical primary HTS data. The first compares an expected assay hit rate to actual hit rates. The second examines the propensity of an individual compound to hit multiple assays. Statistical evaluation of the data indicates a correlation between the resultant functional group filters and compound promiscuity. These data corroborate a number of commonly applied filters as well as producing some unexpected results. Application of these models to HTS collection triage reduced the number of in-house compounds considered for screening by 12%. The implications of these findings are further discussed in the context of the HTS screening set and combinatorial library design as well as compound acquisition.

Measuring CAMD technique performance: a virtual screening case study in the design of validation experiments

The dynamic nature and comparatively young age of computational chemistry is such that novel algorithms continue to be developed at a rapid pace. Such efforts are often wrought at the expense of extensive experimental validations of said techniques, preventing a deeper understanding of their potential utility and limitations. Here we address this issue for ligand-based virtual screening descriptors through design of validation experiments that better reflect the aims of real world application. Applying the newly defined chemotype enrichment approach, a variety of two- and three-dimensional (2D/3D) similarity descriptors have been compared extensively across data sets from four diverse target types. The inhibitors within said data sets contain molecules exhibiting a wide array of substructure functionality, size and flexibility, permitting descriptor comparison in myriad settings. Relative descriptor performance under these conditions is examined, including results obtained using more typical virtual screening validation experiments. Guidelines for optimal application of said descriptors are also discussed in the context of the results obtained, as is the potential utility of fingerprint filtering.

Descriptors you can count on? Normalized and filtered pharmacophore descriptors for virtual screening

The three-dimensional (3D) binary pharmacophore fingerprints find wide application as descriptors in applications ranging from virtual screening through library design. While the 3D content they capture is an intuitively attractive feature of such measures, maximizing their signal to noise ratio has proven to be a tricky balancing act. This issue surfaces primarily due to the potential of such fingerprints to create an explosion of pharmacophores as molecular complexity and flexibility increases. In this article, we describe a modification to the fingerprint generation process that normalizes pharmacophore occurrence frequency by the conformational ensemble size used to derive the descriptor. By including pharmacophore frequency and conformational count, the importance of a given pharmacophore is weighted by the probability of its existence within a given conformational ensemble, rather than treating each pharmacophore equally. In addition, a number of filters have been added to permit the removal of unwanted pharmacophores from the descriptor set. These filters are based on pharmacophore composition (e.g. permutations made up primarily of lipophilic and/or aromatic centers), and size (pharmacophore perimeter length relative to the largest perimeter length found in the molecule). The highly uneven nature of pharmacophore distributions across the conformational ensemble used to generate them is highlighted, as are enrichment comparisons with their binary fingerprint peers. In addition, the limitations in descriptor comparison validation are high-lighted as an illustration of the need for more extensive validation experiments.

Analysis and optimization of structure-based virtual screening protocols. 2. Examination of docked ligand orientation sampling methodology: mapping a pharmacophore for success

An important element of any structure-based virtual screening (SVS) technique is the method used to orient the ligands in the target active site. This has been a somewhat overlooked issue in recent SVS validation studies, with the assumption being made that the performance of an algorithm for a given set of orientation sampling settings will be representative for the general behavior of said technique. Here, we analyze five different SVS targets using a variety of sampling paradigms within the DOCK, GOLD and PROMETHEUS programs over a data set of approximately 10,000 noise compounds, combined with data sets containing multiple active compounds. These sets have been broken down by chemotype, with chemotype hit rate used to provide a measure of enrichment with a potentially improved relevance to real world SVS experiments. The variability in enrichment results produced by different sampling paradigms is illustrated, as is the utility of using pharmacophores to constrain sampling to regions that reflect known structural biology. The difference in results when comparing chemotype with compound hit rates is also highlighted.

Analysis and optimization of structure-based virtual screening protocols (1): exploration of ligand conformational sampling techniques

Ligand conformational flexibility has long been recognized as an important issue in virtual screening (VS). To this end, a number of different methodologies have been adapted to tackle the problem. Many of said techniques were originally designed for ligand derived pharmacophore screens, but have subsequently been fashioned for application within structure-based virtual screening (SVS). A popular adaptation is the pre-calculation of diverse ligand conformations for subsequent docking in target active sites. In this paper, we study a number of the software programs currently being used in conformer generation, analyzing their ability to regenerate known ligand binding conformations. The implications of these studies are discussed, from the perspective of VS in general and SVS in particular.

Analysis and optimization of structure-based virtual screening protocols. (3). New methods and old problems in scoring function design

Scoring function research remains a primary focus of current structure-based virtual screening (SVS) technology development. Here, we present an alternative method for scoring function design that attempts to combine crystallographic structural information with data derived from directly within SVS calculations. The technique utilizes a genetic algorithm (GA) to optimize functions based on binding property data derived from multiple virtual screening calculations. These calculations are undertaken on protein data bank (PDB) complex active sites using ligands of known binding mode in conjunction with "noise" compounds. The advantages of such an approach are that the function does not rely on assay data and that it can potentially use the "noise" binding data to recognize the sub-optimal docking interactions inherent in SVS calculations. Initial efforts in technique exploration using DOCK are presented, with comparisons made to existing DOCK scoring functions. An analysis of the problems inherent to scoring function development is also made, including issues in dataset creation and limitations in descriptor utility when viewed from the perspective of docking mode resolution. The future directions such studies might take are also discussed in detail.

Pyrrolidine-5,5-trans-lactams. 2. The use of X-ray crystal structure data in the optimization of P3 and P4 substituents

[reaction: see text] In this, the second of two letters, we describe the elaboration of the pyrrolidine-5,5-trans-lactam template to delineate the requirements for optimal substitution of the pyrrolidine and lactam nitrogen atoms. Central to the strategy is the use of rapid iterative synthesis in conjunction with X-ray crystal structure determination of ligand-protein complexes.

3-D pharmacophores in drug discovery

In this chapter we review the use of 3-D pharmacophores in drug discovery. Recent advances are highlighted, including the application of pharmacophore descriptors generated both from ligands and protein binding sites. The application of 3-D pharmacophore fingerprints as molecular descriptors for similarity and diversity applications such as virtual screening, library design and QSAR is discussed. In addition, we highlight the quantification of structure-based diversity using site-derived fingerprints, and review virtual screening methods using both single refined hypotheses and the fingerprints of multiple potential hypotheses. Further, we discuss methods that take protein flexibility and molecular shape-into account. Each of the above techniques are reviewed with particular reference to the recent advances, advantages and challenges of each methodology.

Where are the GaPs? A rational approach to monomer acquisition and selection

Gridding and partitioning (GaP) is a computational method for the classification and selection of monomers for combinatorial libraries. The molecules are described in terms of the pharmacophoric groups they contain and where those pharmacophoric groups can be located in three-dimensional space. The approach involves a detailed conformational analysis of each molecule. This conformational analysis is done within a common coordinate frame, thus enabling the monomers to be compared. The use of a partitioned space is central to this particular application as it facilitates the identification of regions of space which are not well represented by existing compounds. Several ways to extend the use of partitioned pharmacophore spaces are described. Applications of the approach in monomer acquisition and in library design are outlined.

New methodology for profiling combinatorial libraries and screening sets: cleaning up the design process with HARPick

Combinatorial chemistry is a tool of increasing importance in the field of ligand design, as it can yield huge increases in the number of compounds available for screening. Unfortunately, it is often the case that the number of molecules which could theoretically be constructed greatly exceeds potential synthesis and screening capacity. For this new technology to be fully exploited, it will become vital to design libraries with reference to the properties of compounds already in existence, if the added value of each new molecular collection is truly to be maximized. Similarly, if we are to take full advantage of the potential of combinatorial chemistry in lead optimization, it is important that our library design paradigms are flexible, with diversity scoring functions that can be modified to suit particular projects. Here these challenges are addressed through the introduction of a novel computer-aided library design tool known as HARPick (heuristic algorithm for reagent picking). The program is accessible to the bench chemist, and incorporates several significant advances over currently available approaches. These include product-based diversity calculations that can be constrained at the reagent level; diversity measures constructed from multiple descriptors; improved pharmacophore key information and full pharmacophore profiling of entire molecular databases. The potential of these improvements to aid in diversity profiling is illustrated through comparison with established methodology, and possible further enhancements are discussed.

Engineering human immunodeficiency virus 1 protease heterodimers as macromolecular inhibitors of viral maturation

Dimerization of human immunodeficiency virus type 1 protease (HIV-1 PR) monomers is an essential prerequisite for viral proteolytic activity and the subsequent generation of infectious virus particles. Disruption of the dimer interface inhibits this activity as does formation of heterodimers between wild-type and defective monomers. A structure-based approach was used to identify amino acid substitutions at the dimer interface of HIV-1 PR that facilitate preferential association of heterodimers and inhibit self-association of the defective monomers. Expression of the designed PR monomers inhibits activity of wild-type HIV-1 PR and viral infectivity when assayed in an ex vivo model system. These results show that it is possible to design PR monomers as macromolecular inhibitors that may provide an alternative to small molecule inhibitors for the treatment of HIV infection.

Molecular docking towards drug discovery

Fueled by advances in molecular structure determination, tools for structure-based drug design are proliferating rapidly. Lead discovery through searching of ligand databases with molecular docking techniques represents an attractive alternative to high-throughout random screening. The size of commercial databases imposes severe computational constraints on molecular docking, compromising the level of calculational detail permitted for each putative ligand. We describe alternative philosophies for docking which effectively address this challenge. With respect to the dynamic aspects of molecular recognition, these strategies lie along a spectrum of models bounded by the Lock-and-Key and Induced-Fit theories for ligand binding. We explore the potential of a rigid model in exploiting species specificity and of a tolerant model in predicting absolute ligand binding affinity. Current molecular docking methods are limited primarily by their ability to rank docked complexes; we therefore place particular emphasis on this aspect of the problem throughout our validation of docking strategies.

Investigating the extension of pairwise distance pharmacophore measures to triplet-based descriptors

Distances between key functional groups have been used for some time as molecular descriptors in 3D database screening and clustering calculations. More recently, a number of groups have explored triplets of molecular centers to describe key ligand features in terms of the properties of triangles. Three-body distances are attractive, since they retain more information than pairwise representations. In most applications, the triangular descriptors have been used to detail molecular shape, using all the constituent atoms or molecular surface points as descriptor centers. As a consequence, the database keying times were such that only single conformers could be considered during molecular descriptor calculations. In this paper we reduce the points used in the molecular description down to the key functional centers, as applied in 3D pharmacophore database searches. Molecular triplets can then be calculated which describe the relative dispositions of differing functional groups, made up from multiple molecular conformations of a given molecule. The new triplet descriptors are compared with classical pairwise distance measures using a variety of pharmacophores, and their potential in database screening, clustering and pharmacophore identification is discussed.

New molecular shape descriptors: application in database screening

Geometric descriptors are becoming popular tools for encoding molecular shape, for use in database screening and clustering calculations. They provide condensed representations of complex objects and, as a consequence, can usually be compared quite rapidly. Here we present a number of new descriptors and methods for the quantification of molecular shape similarity. The techniques are tested using two different biological systems, with particular emphasis on their potential utility as methods for prescreening shape-based database searches. Results are compared with data sets produced using the DOCK program. We find that such similarity evaluations are useful for finding molecules with complementary shape, and that they contain an enriched number of potential DOCK hits when compared to the original databases. Significant limitations in the utility of such DOCK prescreens are discussed, and potential solutions are considered.

QSAR's from similarity matrices. Technique validation and application in the comparison of different similarity evaluation methods

It has recently been shown that good quantitative structure-activity relationships can be obtained through statistical analysis of molecular similarity matrices. Here we extend the technique to seven additional molecular series, previously studied using Comparative Molecular Field Analysis (CoMFA) methodology. The results are used to confirm technique applicability across a wider range of QSAR problems and to compare quantitatively the ability of various similarity indices to describe biological systems. The relative merits of this technique in comparison to CoMFA are discussed.

Structure-activity relationships from molecular similarity matrices

An alternative method for determining structure-activity correlations is presented. Ligand molecules are described using data matrices derived from the results of N by N (each molecule compared to every other) molecular similarity calculations. The matrices were analyzed using a neural network pattern recognition technique and partial least squares statistics, with the results obtained compared to those achieved using comparative molecular field analysis (CoMFA). The molecular series used in the study comprised 31 steroids. The resultant pattern recognition analysis showed clustering of compounds with high, intermediate, and low affinity into separate regions of the neuron output plots. The cross-validated correlation coefficients obtained from statistical analyses of the matrices against steroid binding data compared well with those achieved using CoMFA. These results show that data matrices derived from molecular similarity calculations can provide the basis for rapid elucidation of both qualitative and quantitative structure-activity relationships.

Similarity screening of molecular data sets

Three-dimensional (3D)-database searches are now being widely applied to determine potential new active molecules. Many structural data sets obtained as a result of these searches are still large in size. In this paper we apply molecular similarity calculations as a rapid method to screen two such data sets. In the first investigation, synthetic candidates, produced as a result of a tendamistat beta-turn mimic search, were tested for their ability to imitate the beta-turn backbone. In the second study, structures extracted through a histamine pharmacophore query search were examined on the basis of their electronic similarity to histamine. Molecular similarity is shown to provide a rapid means of gaining insight into the composition of molecular data sets, with possible implications for future full 3D-database searches.

The calculation of molecular similarity: alternative formulas, data manipulation and graphical display

The use of electrostatic potential comparisons between molecules for the elucidation of structure activity relationships is now a well-established modeling technique. The Carbo and Hodgkin similarity indices are used extensively to make quantitative comparisons of this nature; yet their roots are found in the overlap of electron density distribution, with both formulas utilizing a product-based numerator. Two new similarity indices are suggested that calculate the electrostatic potential similarity using a difference-based numerator. The form of the new indices allows the creation of additional software functions that enhance the flexibility of similarity calculations and permit the creation of similarity maps. The general properties of these software functions and all indices are discussed and applied to a series of dopamine D2 receptor agonists.