Search

Merck & Co., Inc., Kenilworth, NJ, USA, is undergoing a transformation in the way that it prosecutes R&D programs. Through the adoption of a “model-driven” culture, enhanced R&D productivity is anticipated, both in the form of decreased attrition at each stage of the process and by providing a rational framework for understanding and learning from the data generated along the way. This new approach focuses on the concept of a “Design Cycle” that makes use of all the data possible, internally and externally, to drive decision-making. These data can take the form of bioactivity, 3D structures, genomics, pathway, PK/PD, safety data, etc. Synthesis of high-quality data into models utilizing both well-established and cutting-edge methods has been shown to yield high confidence predictions to prioritize decision-making and efficiently reposition resources within R&D. The goal is to design an adaptive research operating plan that uses both modeled data and experiments, rather than just testing, to drive project decision-making. To support this emerging culture, an ambitious information management (IT) program has been initiated to implement a harmonized platform to facilitate the construction of cross-domain workflows to enable data-driven decision-making and the construction and validation of predictive models. These goals are achieved through depositing model-ready data, agile persona-driven access to data, a unified cross-domain predictive model lifecycle management platform, and support for flexible scientist-developed workflows that simplify data manipulation and consume model services. The end-to-end nature of the platform, in turn, not only supports but also drives the culture change by enabling scientists to apply predictive sciences throughout their work and over the lifetime of a project. This shift in mindset for both scientists and IT was driven by an early impactful demonstration of the potential benefits of the platform, in which expert-level early discovery predictive models were made available from familiar desktop tools, such as ChemDraw. This was built using a workflow-driven service-oriented architecture (SOA) on top of the rigorous registration of all underlying model entities.

The development of new antimalarial therapies is essential, and lowering the barrier of entry for the screening and discovery of new lead compound classes can spur drug development at organizations that may not have large compound screening libraries or resources to conduct high-throughput screens. Machine learning models have been long established to be more robust and have a larger domain of applicability with larger training sets. Screens over multiple data sets to find compounds with potential malaria blood stage inhibitory activity have been used to generate multiple Bayesian models. Here we describe a method by which Bayesian quantitative structure-activity relationship models, which contain information on thousands to millions of proprietary compounds, can be shared between collaborators at both for-profit and not-for-profit institutions. This model-sharing paradigm allows for the development of consensus models that have increased predictive power over any single model and yet does not reveal the identity of any compounds in the training sets.

Fragment based screening (FBS) has emerged as a mainstream lead discovery strategy in academia, biotechnology start-ups, and large pharma. As a prerequisite of FBS, a structurally diverse library of fragments is desirable in order to identify chemical matter that will interact with the range of diverse target classes that are prosecuted in contemporary screening campaigns. In addition, it is also desirable to offer synthetically amenable starting points to increase the probability of a successful fragment evolution through medicinal chemistry. Herein we describe a method to identify biologically relevant chemical substructures that are missing from an existing fragment library (chemical gaps), and organize these chemical gaps hierarchically so that medicinal chemists can efficiently navigate the prioritized chemical space and subsequently select purchasable fragments for inclusion in an enhanced fragment library.

Fragment based screening (FBS) has emerged as a mainstream lead discovery strategy in academia, biotechnology start-ups, and large pharma. As a prerequisite of FBS, a structurally diverse library of fragments is desirable in order to identify chemical matter that will interact with the range of diverse target classes that are prosecuted in contemporary screening campaigns. In addition, it is also desirable to offer synthetically amenable starting points to increase the probability of a successful fragment evolution through medicinal chemistry. Herein we describe a method to identify biologically relevant chemical substructures that are missing from an existing fragment library (chemical gaps), and organize these chemical gaps hierarchically so that medicinal chemists can efficiently navigate the prioritized chemical space and subsequently select purchasable fragments for inclusion in an enhanced fragment library.

Understanding the development of a scientific approach is a valuable exercise in gauging the potential directions the process could take in the future. The relatively short history of applying computational methods to absorption, distribution, metabolism and excretion (ADME) can be split into defined periods. The first began in the 1960s and continued through the 1970s with the work of Corwin Hansch et al. Their models utilized small sets of in vivo ADME data. The second era from the 1980s through 1990s witnessed the widespread incorporation of in vitro approaches as surrogates of in vivo ADME studies. These approaches fostered the initiation and increase in interpretable computational ADME models available in the literature. The third era is the present were there are many literature data sets derived from in vitro data for absorption, drug–drug interactions (DDI), drug transporters and efflux pumps [P-glycoprotein (P-gp), MRP], intrinsic clearance and brain penetration, which can theoretically be used to predict the situation in vivo in humans. Combinatorial synthesis, high throughput screening and computational approaches have emerged as a result of continual pressure on pharmaceutical companies to accelerate drug discovery while decreasing drug development costs. The goal has become to reduce the drop-out rate of drug candidates in the latter, most expensive stages of drug development. This is accomplished by increasing the failure rate of candidate compounds in the preclinical stages and increasing the speed of nomination of likely clinical candidates. The industry now understands the reasons for clinical failure other than efficacy are mainly related to pharmacokinetics and toxicity. The late 1990s saw significant company investment in ADME and drug safety departments to assess properties such as metabolic stability, cytochrome P-450 inhibition, absorption and genotoxicity earlier in the drug discovery paradigm. The next logical step in this process is the evaluation of higher throughput data to determine if computational (in silico) models can be constructed and validated from it. Such models would allow an exponential increase in the number of compounds screened virtually for ADME parameters. A number of researchers have started to utilize in silico, in vitro and in vivo approaches in parallel to address intestinal permeability and cytochrome P-450-mediated DDI. This review will assess how computational approaches for ADME parameters have evolved and how they are likely to progress.

The potential of ostensibly structurally diverse environmental chemicals to modulate endocrine processes in biological systems has been recognized. Difficulty in classifying endocrine system modulators by chemical structure may in large part be due to lack of understanding of mechanisms of action. New developments in understanding nuclear receptor mechanisms of hormone action support a more complex mechanism, possibly involving dimerization/aggregation events leading to multimeric receptor complexes in agonist action. Because of the requirement for high structural specificity in agonist action, it is suggested that most environmental chemicals of concern are likely to function as imperfect hormones with partial agonist-antagonist properties, especially at environmentally realistic concentrations. In the absence of having appropriately placed molecular recognition domains to affect agonist action, partial agonism-antagonism may be associated with favorable low-energy conformational flexibility and complementary receptor protein flexibility. The halogenated aromatic hydrocarbons are of particular concern as hormone mimics since they often have (1) similar molecular recognition factors but in many cases relatively more flexible structures, (2) similar bulk physico-chemical properties controlling uptake and distribution in biological systems, and (3) are relatively more resistant to metabolism and elimination. Some important molecular reactivity properties underlying thyromimetic and estrogenic actions of some of these chemicals are identified and described in terms of structure-activity relationships (SARs). It is proposed that specificity of hormone action in the nucleus could be associated with differential interaction of ligand-bound receptor dimeric forms with other transcription factors specific to the target cell. The small-molecule ligand can be viewed as playing a central, multifunctional role in nuclear receptor action as an organic unmasking and reclustering agent for critical macromolecules. Evidence is discussed in support of a nuclear heterodimerization model for dioxin and related compound action involving a structural transition mechanism. These models with some molecular detail also have utility in understanding the different structural properties of agonists and antagonists. There would appear to be ample opportunities for environmental chemicals to act as antagonists for multiple receptor systems with little more than anchor-ring similarities in structure. The application of three-dimensional quantitative structure-activity (3D QSAR) models incorporating such structural information should be a useful adjunct for identifying endocrine system modulating chemicals. This data has implications for (1) improved drug design, (2) understanding of chemical interaction toxicity, (3) removing undesirable chemicals from our environment, and (4) reducing their chemical release.

Our previous reports indicate that ortho-substituted non-coplanar polychlorinated biphenyl (PCB) congeners perturbed neuronal Ca2+-homeostasis in vitro, altered agonist-stimulated inositol phosphate accumulation, and caused protein kinase C (PKC) translocation. The structure-activity relationship (SAR) with 24 PCB congeners was consistent with a chlorination pattern that favored non-coplanarity while those with chlorination that favored coplanarity were less active. To test the hypothesis that coplanarity (or lack thereof) is a significant factor in the activity of PCBs, studies with related classes of chemicals such as the polychlorinated diphenyl ethers (PCDEs), in which coplanarity is more difficult to achieve regardless of degree and pattern of chlorination, were initiated. The selected PCDEs and their analogs are predicted to be active, since they are non-coplanar in nature. The effects of these chemicals were studied using the same measures for which PCBs had differential effects based on structural configuration. These measures include PKC translocation as determined by [3H]-phorbol ester ([3H]PDBu) binding in cerebellar granule cells and 45Ca2+ sequestration as determined by 45Ca2+ uptake by microsomes and mitochondria isolated from adult rat cerebellum. All the PCDE congeners studied, increased [3H]PDBu binding in a concentration-dependent manner. The order of potency was 2,4,4'-trichlorodiphenyl ether > 4,4'-dichlorodiphenyl ether > diphenyl ether, 3,3',4,4'-tetrachlorodiphenyl ether and, 2,2',4,4',5- and 2,3',4,4',5-pentachlorodiphenyl ethers. The structurally related diphenyl ether nitrofen and diphenyl ethanes o,p'-1,1,1-trichloro-2,2-bis[p-chlorophenyl]ethane (DDT) and p,p'-DDT increased [3H]PDBu binding to a similar extent (28-35% stimulation at 100 microM). All PCDE congeners and their analogs inhibited 45Ca2+ sequestration by microsomes and mitochondria. Of all the chemicals, unchlorinated diphenyl ether was the least active. These results are in agreement with previous SAR findings in which non-coplanar PCBs are active and support our hypothesis that the extent of coplanarity determined by a pattern of chlorination on certain aromatic hydrocarbons can weaken their potency in vitro, although the extent of chlorination is also important.

VALIDATE is a hybrid approach to predict the binding affinity of novel ligands for receptors of known three-dimensional structure. This approach calculates physicochemical properties of the ligand and the receptor - ligand complex to estimate the free energy of binding. The enthalpy of binding is calculated by molecular mechanics while properties such as complementary hydrophobic surface area are used to estimate the entropy of binding through heuristics. A diverse training set of 51 crystalline complexes was assembled, and their relevant physicochemical properties were computed. These properties were analyzed by partial least squares (PLS) statistics, or neural network analysis (SONNIC), to generate models for the general prediction of the affinity of ligands with receptors of known three-dimensional structure. The ability of the model to predict the affinity of novel complexes not included in the training set was demonstrated with three independent test sets: 14 complexes of known three-dimensional structure including 3 DNA complexes, a class of compound not included in the training set, 13 HIV protease inhibitors fit to HIV-1 protease, and 11 thermolysin inhibitors fit to thermolysin.

Comparative molecular field analysis (CoMFA), a three-dimensional quantitative structure-activity relationship (3D-QSAR) paradigm, was used to examine the estrogen receptor (ER) binding affinities of a series of structurally diverse natural, synthetic, and environmental chemicals of interest. The CoMFA/3D-QSAR model is statistically robust and internally consistent, and successfully illustrates that the overall steric and electrostatic properties of structurally diverse ligands for the estrogen receptor are both necessary and sufficient to describe the binding affinity. The ability of the model to accurately predict the ER binding affinity of an external test set of molecules suggests that structure-based 3D-QSAR models may be used to supplement the process of endocrine disrupter identification through prioritization of novel compounds for bioassay. The general application of this 3D-QSAR model within a toxicological framework is, at present, limited only by the quantity and quality of biological data for relevant biomarkers of toxicity and hormonal responsiveness. 28 refs., 12 figs., 9 tabs.

Drug discovery is shifting focus from industry to outside partners and, in the process, creating new bottlenecks. Technologies like high throughput screening (HTS) have moved to a larger number of academic and institutional laboratories in the USA, with little coordination or consideration of the outputs and creating a translational gap. Although there have been collaborative public–private partnerships in Europe to share pharmaceutical data, the USA has seemingly lagged behind and this may hold it back. Sharing precompetitive data and models may accelerate discovery across the board, while finding the best collaborators, mining social media and mobile approaches to open drug discovery should be evaluated in our efforts to remove drug discovery bottlenecks. We describe four strategies to rectify the current unsustainable situation.

Drug Discovery is shifting focus from the industry to outside partners and in the process creating new bottlenecks, suggesting the need for a more disruptive overhaul. Technologies like high throughput screening (HTS) have moved to a larger number of academic and institutional laboratories in the US, with little apparent coordination or consideration of the outputs and creating a translational gap. While there have been collaborative public private partnerships in Europe to share pharmaceutical data, the USA has lagged behind. Sharing precompetitive computational models may be the next frontier to provide more confidence in the quality of the leads produced and attract investment. We suggest there needs to be an awareness of what research is going on in the screening centers, more collaboration and coordination. These efforts will shift the focus to finding the best researchers to fund and require a rethink of how to reward their collaborative efforts.

NewPred, a semiautomated procedure to evaluate alternate binding modes and assist three dimensional quantitative structure-activity relationship (3D-QSAR) studies in predictive power evaluation is exemplified with a series of 30 human immunodeficiency virus 1 protease (HIV PR) inhibitors. Five comparative molecular field analysis (CoMFA) models (Waller, C. L.; et al. J. Med. Chem. 1993, 36, 4152-4160) based on 59 HIV-PR inhibitors were tested. The test set included 18 compounds (set A) having a different transition state isostere (TSI), hydroxyethylurea (Getman, D. P.; et al. J. Med. Chem. 1993, 36, 288-291), to investigate the binding mode in P1' and P2'. Twelve dihyroxyethylenes (set B) (Thaisrivongs, S.; et al. J. Med. Chem. 1993, 36, 941-952) were used to investigate binding in P2 and P3 as well as in P2' and P3'. Six other compounds with known or inferred binding structure (set C) were part of the test set, but not investigated with NewPred. Each compound was aligned in accordance to predefined alignment rules for the training set prior to the inclusion in the test set (except for set C). Using NewPred, geometrically different conformers for each compound were generated and individually relaxed in the HIV-PR binding site. Energy comparisons allowed selection of lowest energy structures to be included in the test set. Only in vacuo minimized conformers derived from low-energy complexes were used to determine the predictive power of the five models (predictive r2 varied from 0.1 to 0.7 when two chemical and statistical outliers were excluded). Our models correctly predict the poor inhibitor activity of 1(S)-amino-2(R)-hydroxyindan-containing peptides (set B), which is explained and interpreted from a 3D-QSAR perspective. The use of a new, flexibility-based, semiautomated method to explore alternate binding models for 3D-QSAR models is demonstrated.

It has been demonstrated that agents which inhibit chloride influx and, therefore, lower intracellular chloride levels in the astrocyte, a major cell type in the cerebral gray matter, inhibit astrocytic swelling in vitro and in vivo . Herein, we report additional examples of a series of [( N -alkyl-1,3-dihydro-1-oxoisoindolin-5-yl)oxy] alkanoic acids and their effects upon ion transport in primary rat astrocyte cultures. The 4-chloro-substituted 1-oxoisoindolines demonstrated superior astrocytic chloride influx inhibitory activity as compared to the 6-chloro and non-chlorinated analogs. The four-carbon acid side chain derivatives were more active than the three- and two-carbon analogs. The pharmacological profile of these compounds was examined with respect to inhibition of the Cl − -Cl − /Cl − -HCO 3 − anion exchanger and Na + -K + -2Cl − cotransport mechanisms in glia, and the compounds were found to exhibit a similar profile to that of furosemide by inhibiting both transporters.

Molecular modeling studies were carried out on a series of 1-oxoisoindolines which are pharmacologically active as inhibitors of astrocytic chloride transport. Conformational analysis revealed that the halogen substituent exerted a pronounced steric directing effect on the acid side chain. The 4-substituted analogs apparently provided for the best spatial arrangement of pharamacophoric elements of the molecules. Conventional quantitative structure-activity relationship (QSAR) studies using lipophilic and dipole moment characteristics of the molecules as physical descriptor variables in the regression equation yielded a statistically significant model. Comparative molecular field analysis (CoMFA) was utilized as a three-dimensional QSAR technique to explore changes in the steric and electrostatic fields of the molecules that can account for differences in biological activity values. A highly predictive model was attained which supported the results from the qualitative and conventional quantitative structure-activity relationship analyses. These modeling techniques represent the evolutionary process by which structure-activity methods were employed to aid in the development of novel more potent inhibitors of astrocytic chloride transport.

Octanol-water partition coefficients, as logPs, are frequently used to describe transportability and membrane permeability of chemical compounds. Because of this they represent an integral part of quantitative structure-activity relationship (QSAR) theory. Since the analytical determination of logPs is a tedious process, many empirical techniques for their calculation have been developed. The additive fragment-based technique, clogP, represents the most highly validated algorithm designed for this purpose. However, it has been observed that clogP is incapable of distinguishing between structural isomers in a congeneric series of compounds. This is primarily a result of the lack of necessary “factors” for the characterization of the substituent effects. As an alternative to additional parameterization of the clogP methodology, a novel technique for the calculation of octanol-water partition coefficients based on three-dimensional descriptors was developed which is capable of accurately predicting the experimentally-determined logPs for a congeneric series of structural isomers.

The utility of comparative molecular field analysis (CoMFA), a three-dimensional Quantitative Structure-Activity Relationship (3-D QSAR) paradigm, as a tool to aid in the development of predictive models has been previously addressed (Depriest, S.D. et al., J. Am. Chem. Soc. 1993, in press). Although predictive correlations were obtained for angiotensin-converting and thermolysin inhibitors, certain inadequacies of the CoMFA technique were noted. Primarily, CoMFA steric and electrostatic fields alone do not fully characterize the zinc-ligand interaction. Previously, this was partially rectified by the inclusion of indicator variables into the QSAR table to designate the class of zinc-binding ligand. Recent advances in molecular modeling technology have allowed us to further address this limitation of the preceding study. Using molecular orbital fields derived from semiempirical calculations as additional descriptors in the QSAR table, predictive correlations were produced based on CoMFA and molecular orbital fields alone--indicator variables no longer being necessary. Arbitrary information concerning the alignment of molecules under study within the active-site introduces ambiguities into the CoMFA study. Crystallographic information detailing the binding mode of several thermolysin enzyme inhibitors has previously been used as a guide for the alignment of additional, noncrystallized, inhibitors. However, this process was complicated by the lack of parameters for zinc in the molecular mechanical force field. Therefore, zinc-ligand interactions were ignored during the standard minimization procedure. The use of field-fit minimization using complementary receptor fields as templates is presented as a possible solution to the problem. Predictive correlations were obtained from analyses based on this method of molecular alignment. The availability of crystallographic data for thermolysin enzyme-inhibitor complexes allowed for an alternate definition of the CoMFA region. Herein, promising results from analyses using actual receptor active-site atom probe atoms are presented.

PGVL Hub is an integrated molecular design desktop tool that has been developed and globally deployed throughout Pfizer discovery research units to streamline the design and synthesis of combinatorial libraries and singleton compounds. This tool supports various workflows for design of singletons, combinatorial libraries, and Markush exemplification. It also leverages the proprietary PGVL virtual space (which contains 10(14) molecules spanned by experimentally derived synthesis protocols and suitable reactants) for lead idea generation, lead hopping, and library design. There had been an intense focus on ease of use, good performance and robustness, and synergy with existing desktop tools such as ISIS/Draw and SpotFire. In this chapter we describe the three-tier enterprise software architecture, key data structures that enable a wide variety of design scenarios and workflows, major technical challenges encountered and solved, and lessons learned during its development and deployment throughout its production cycles. In addition, PGVL Hub represents an extendable and enabling platform to support future innovations in library and singleton compound design while being a proven channel to deliver those innovations to medicinal chemists on a global scale.

Ligand-based computational models could be more readily shared between researchers and organizations if they were generated with open source molecular descriptors [e.g., chemistry development kit (CDK)] and modeling algorithms, because this would negate the requirement for proprietary commercial software. We initially evaluated open source descriptors and model building algorithms using a training set of approximately 50,000 molecules and a test set of approximately 25,000 molecules with human liver microsomal metabolic stability data. A C5.0 decision tree model demonstrated that CDK descriptors together with a set of Smiles Arbitrary Target Specification (SMARTS) keys had good statistics [κ = 0.43, sensitivity = 0.57, specificity = 0.91, and positive predicted value (PPV) = 0.64], equivalent to those of models built with commercial Molecular Operating Environment 2D (MOE2D) and the same set of SMARTS keys (κ = 0.43, sensitivity = 0.58, specificity = 0.91, and PPV = 0.63). Extending the dataset to ∼193,000 molecules and generating a continuous model using Cubist with a combination of CDK and SMARTS keys or MOE2D and SMARTS keys confirmed this observation. When the continuous predictions and actual values were binned to get a categorical score we observed a similar κ statistic (0.42). The same combination of descriptor set and modeling method was applied to passive permeability and P-glycoprotein efflux data with similar model testing statistics. In summary, open source tools demonstrated predictive results comparable to those of commercial software with attendant cost savings. We discuss the advantages and disadvantages of open source descriptors and the opportunity for their use as a tool for organizations to share data precompetitively, avoiding repetition and assisting drug discovery.

Variable selection is typically a time-consuming and ambiguous procedure in performing quantitative structure−activity relationship (QSAR) studies on overdetermined (regressor-heavy) data sets. A variety of techniques including stepwise and partial least squares/principlal components analysis (PLS/PCA) regression have been applied to this common problem. Other strategies, such as neural networks, cluster significance analysis, nearest neighbor, or genetic (function) or evolutionary algorithms have also evaluated. A simple random selection strategy that implements iterative generation of models, but directly avoids crossover and mutation, has been developed and is implemented herein to rapidly identify from a pool of allowable variables those which are most closely associated with a given response variable. The FRED (fast random elimination of descriptors) algorithm begins with a population of offspring models composed of either a fixed or variable number of randomly selected variables. Iterative eliminati...

Comparative molecular field analysis (CoMFA), a three-dimensional quantitative structure−activity relationship (3D-QSAR) technique, has proven to be a valuable tool in the field of rational drug design. In its native form, CoMFA utilizes a pseudoreceptor, in the form of a regularly spaced lattice of probe atoms, to characterize the steric and electrostatic properties of a series of mutually superimposed molecules. Statistical analyses are performed in an attempt to correlate changes in these shape and charge related fields to observed differences in biological activities at a given target. Graphical analyses of the resulting “negative receptor images” have been demonstrated to provide insight into the physicochemical requirements of novel ligands. Several groups have previously demonstrated the benefits of additional or alternative fields for these types of analyses. In this report, a novel molecular potential field derived from atomistic contributions to molecular polarizability is presented. Comparison ...

Cheminformatics is at a turning point, the pharmaceutical industry benefits from using the various methods developed over the last twenty years, but in our opinion we need to see greater development of novel approaches that non-experts can use. This will be achieved by more collaborations between software companies, academics and the evolving pharmaceutical industry. We suggest that cheminformatics should also be looking to other industries that use high performance computing technologies for inspiration. We describe the needs and opportunities which may benefit from the development of open cheminformatics technologies, mobile computing, the movement of software to the cloud and precompetitive initiatives.

Between 2004 and 2008, the US National Institutes of Health Molecular Libraries and Imaging initiative pilot phase funded 10 high-throughput screening centers, resulting in the deposition of 691 assays into PubChem and the nomination of 64 chemical probes. We crowdsourced the Molecular Libraries and Imaging initiative output to 11 experts, who expressed medium or high levels of confidence in 48 of these 64 probes.

Much progress has been made over the past several years to provide technologies for the integration of drug discovery software applications and the underlying data bits. Integration at the application layer has focused primarily on developing and delivering applications that support specific workflows within the drug discovery arena. A fine balance between creating behemoth applications and providing business value must be maintained. Heterogeneous data sources have typically been integrated at the data level in an effort to provide a more holistic view of the data packages supporting key decision points. This review will highlight past attempts, current status, and potential future directions for systems and data integration strategies in support of drug discovery efforts.

Preventing drug-drug interactions and reducing drug-related mortalities dictate cleaner and costlier medicines. The cost to bring a new drug to market has increased dramatically over the last 10 years, with post-discovery activities (preclinical and clinical) costs representing the majority of the spend. With the ever-increasing scrutiny that new drug candidates undergo in the post-discovery assessment phases, there is increasing pressure on discovery to deliver higher-quality drug candidates. Given that compound attrition in the early clinical stages can often be attributed to metabolic liabilities, it has been of great interest lately to implement predictive measures of metabolic stability/ liability in the drug design stage of discovery. The solution to this issue is wrapped in understanding the basic of the cytochrome P450 (CYP) enzymes functions and structures. Recently, experimental information on the structure of a variety of cytochrome P450 enzymes, major contributors to phase I metabolism, has become readily available. This, coupled with the availability of experimental information on substrate specificities, has lead to the development of numerous computational models (macromolecular, pharmacophore, and structure-activity) for the rationalization and prediction of CYP liabilities. A comprehensive review of these models is presented in this chapter.

Preventing drug-drug interactions and reducing drug-related mortalities dictate cleaner and costlier medicines. The cost to bring a new drug to market has increased dramatically over the last 10 years, with post-discovery activities (preclinical and clinical) costs representing the majority of the spend. With the ever-increasing scrutiny that new drug candidates undergo in the post-discovery assessment phases, there is increasing pressure on discovery to deliver higher-quality drug candidates. Given that compound attrition in the early clinical stages can often be attributed to metabolic liabilities, it has been of great interest lately to implement predictive measures of metabolic stability/ liability in the drug design stage of discovery. The solution to this issue is wrapped in understanding the basic of the cytochrome P450 (CYP) enzymes functions and structures. Recently, experimental information on the structure of a variety of cytochrome P450 enzymes, major contributors to phase I metabolism, has become readily available. This, coupled with the availability of experimental information on substrate specificities, has lead to the development of numerous computational models (macromolecular, pharmacophore, and structure-activity) for the rationalization and prediction of CYP liabilities. A comprehensive review of these models is presented in this chapter.

Both qualitative and quantitative modeling methods relating chemical structure to biological activity, called structure-activity relationship analyses or SAR, are applied to the prediction and characterization of chemical toxicity. This minireview will discuss some generic issues and modeling approaches that are tailored to problems in toxicology. Different approaches to, and some facets and limitations of the practice and science of, SAR as they pertain to current toxicology analyses, and the basic elements of SAR and SAR-model development and prediction systems are discussed. Other topics include application of 3-D SAR to understanding of the propensity of chemicals to cause endocrine disruption, and the use of models to analyze biological activity of metal ions in toxicology. An example of integration of knowledge pertaining to mechanisms into an expert system for prediction of skin sensitization to chemicals is also discussed. This minireview will consider the utility of modeling approaches as one component for better integration of physicochemical and biological properties into risk assessment, and also consider the potential for both environmental and human health effects of chemicals and their interactions.

Variable selection is typically a time-consuming and ambiguous procedure in performing quantitative structure-activity relationship (QSAR) studies on over-determined (regressor-heavy) data sets. A variety of techniques including stepwise and partial least squares/principle components analysis (PLS/PCA) regression have been applied to this common problem. Other strategies, such as neural networks, cluster significance analysis, nearest neighbor, or genetic (function) or evolutionary algorithms have also evaluated. A simple random selection strategy that implements iterative generation of models, but directly avoids cross-over and mutation, has been developed and is implemented herein to rapidly identify from a pool of allowable variables, those which are most closely associated with a given response variable. The FRED (fast random elucidation of determinants) algorithm begins with a population of offspring (models) composed of a fixed, or variable, number of randomly selected variables. Iterative elimination of descriptors leads naturally to subsequent generations of more fit offspring (models). In contrast to common genetic and evolutionary algorithms, only those descriptors determined to contribute to the genetic make-up of less fit offspring (models) are eliminated from the descriptor pool. After every generation, a new random increment line search of the remaining descriptors initiates the development of the next generation of randomly constructed models. An optional algorithm with eliminates highly correlated descriptors in a stepwise manner prior to the development of the first generation of offspring greatly enhances the efficiency of the FRED algorithm. A FRED analysis on a set of antifilarials published by Selwood (n=31 compounds, k=53 descriptors) demonstrates the ability of the algorithm to rapidly identify determinants of biological outcome form a large collection of highly intercorrelated variables (see Figure 1.). A comparison of the results of a FRED analysis of the Selwood data set with those obtained using alternative algorithms reveals that this technique is capable of identifying the same “optimal” solutions in an efficient manner.

Carbon tetrachloride (CCl4) is a potent hepatotoxic agent whose toxicity is mediated through cytochome P450-dependent metabolism. Results from anaerobic in vitro experiments with hepatic microsomes isolated from male F-344 rats indicate that chlorofom (CHCl3) formation from CCl4 is nonlinear with dose. Dose is traditionally expressed as the amount of CCl4 added to the vial. In this study, a pharmacokinetic model has been developed to calculate the concentration of CCl4 in the microsomal suspension. Hepatic microsomes prepared from fed and fasted animals were incubated with CCl4 under anaerobic conditions and formation of CHCl3 over a 5-min incubation period was monitored by headspace gas chromatography. Dose-response curves, based on total amount of CCl4 added to the microsomes, revealed a nonlinear, biphasic appearance of CHCl3, with fasting slightly increasing CHCl3 production in microsomes prepared from fasted rats. Microsomes were also pretreated with the CYP2E1 inhibitor, diallyl sulfone (DAS), before addition of CCl4. In uninhibited microsomes, there appeared to be a high-affinity saturable phase of metabolism occurring at lower concentrations followed by a linear phase at higher CCl4 concentrations. Following DAS pretreatment, the saturable portion of the dose-response curve was inhibited more than the linear phase with the biphasic CHCl3 production becoming more linear. DAS inhibition eliminated the effect of fasting on CHCl3 formation. The best fit kinetic constants for the saturable phase resulted in an estimate of V(max) of 0.017 mg/h/mg protein (V(maxc) = 7.61 mg/h/kg) and Km of 2.3 mg/l (15 microM). The linear phase rate constant (kf) was determined to be 0.046 h-1) (kfc = 0.03 h-1). In conclusion, a pharmacokinetic model has been developed for anaerobic in vitro metabolism of CCl4 to CHCl3 that estimates metabolic rates based on CHCl3 formation and actual CCl4 concentration in the microsomal suspension.

In the present study we have utilized comparative molecular field analysis (CoMFA), a three-dimensional quantitative structure-activity relationship paradigm, to explore the physico-chemical requirements for binding to the Ah (dioxin) receptor. Recent developments by Gillner et al. [(1993) Mol. Pharmacol. 44, 336-345] prompted us to review and revise our previous CoMFA/QSAR model [Waller, C. L., and McKinney, J. D. (1992) J. Med. Chem. 36, 3660-3666] to include a structurally-diverse training set of Ah receptor ligands ranging in size from naphthalene to indolo[3,2-b]carbazole nuclei. An exhaustive validation process utilizing external test sets and hierarchical cluster analysis routines was employed during model construction and is discussed herein. The limitations of the approach presented herein are discussed with respect to predictive ability of the CoMFA/QSAR models, which is demonstrated to be dependent on a balance between structural diversity and redundancy in the molecules comprising the training set. The results of our modified CoMFA/QSAR model are consistent with and unify all previously established structure-activity relationships established for less structurally-diverse training sets of Ah receptor ligands. As a result of the more complete nature of the series of molecules under examination in the present study, the CoMFA/QSAR steric and electrostatic field contour plots as well as the essential and excluded volume plots provide for a more detailed characterization of the molecular binding domain of the Ah receptor. The implications of the CoMFA/QSAR model presented herein are explored with respect to quantitative hazard identification of potential toxicants.

Comparative molecular field analysis (CoMFA), a three-dimensional, quantitative structure-activity relationship (QSAR) paradigm, was used to examine the correlations between the calculated physicochemical properties and the in vitro activities of a series of human immunodeficiency virus (HIV-1) protease inhibitors. The training set consisted of 59 molecules from five structurally-diverse transition-state isostere classes: hydroxyethylamine, statine, norstatine, keto amide, and dihydroxyethylene. The availability of X-ray crystallographic data for at least one representative from each class bound to the protease provided information regarding not only the active conformation of each ligand but also, via superimposition of protease backbones, the relative positions of each ligand with respect to one another in the active site of the enzyme. Once aligned, these molecules served as templates on which additional congeners were field-fit minimized. Additional alignment rules were derived from minimizations of the ligands in the active site of the semirigid protease. The predictive ability of each resultant model was evaluated using a test set comprised of molecules containing a novel transition-state isostere: hydroxyethylurea. Crystallographic studies (Getman, D.P.; et al. J. Med. Chem. 1993, 36, 288-291) indicated an unexpected binding mode for this series of compounds which precluded the use of the field-fit minimization alignment technique. The test set molecules were, therefore, subjected to a limited systematic search in conjunction with active-site minimization. The conformer of each molecule expressing the lowest interaction energy with the active site was included in the test set. Field-fit minimization of neutral molecules to crystal ligands and active-site minimizations of protonated ligands yielded predictive correlations for HIV-1 protease inhibitors. The use of crystallographic data in the determination of alignment rules and field-fit minimization as a molecular alignment tool in the absence of direct experimental data regarding binding modes is strongly supported by these results.

Nat. Chem. Biol. 5, 441–447 (2009); published online 17 June 2009; corrected after print 17 June 2009 In the version of this article initially published, the phrase “Though MLI and its pilot phase budget and superficially modest productivity were subject to industrial criticism” originally cited reference 3, but should cite reference 2.

Certain phenyl-substituted hydrocarbons of environmental concern have the potential to disrupt the endocrine system of animals, apparently in association with their estrogenic properties. Competition with natural estrogens for the estrogen receptor is a possible mechanism by which such effects could occur. We used comparative molecular field analysis (CoMFA), a three-dimensional quantitative structure-activity relationship (QSAR) paradigm, to examine the underlying structural properties of ortho-chlorinated hydroxybiphenyl analogs known to bind to the estrogen receptor. The cross-validated and conventional statistical results indicate a high degree of internal predictability for the molecules included in the training data set. In addition to the phenolic (A) ring system, conformational restriction of the overall structure appears to play an important role in estrogen receptor binding affinity. Hydrophobic character as assessed using hydropathic interaction fields also contributes in a positive way to binding affinity. The CoMFA-derived QSARs may be useful in examining the estrogenic activity of a wider range of phenyl-substituted hydrocarbons of environmental concern. 37 refs., 2 figs., 2 tabs.