Development and experimental validation of a docking strategy for the generation of kinase-targeted libraries.

A high-throughput docking strategy for the filtering of in silico compounds and the generation of kinase-targeted libraries is described. Systematic docking and scoring in three kinase crystal 3D structures of 123 structurally diverse kinase ligands led to the determination of six thresholds for each kinase. These thresholds were used as filters for the virtual screening of two collections of compounds: a collection of more than 2500 drugs and drug-like compounds (negative control) and a kinase-targeted library of 1440 compounds. This strategy was then experimentally validated by testing 60 compounds from the kinase-targeted library on 41 kinases from five different families. The 60 compounds were split into those passing all the thresholds and the others (30 compounds in each group). The overall hit enrichment was 6.70-fold higher in the first group, validating our approach for the generation of kinase-targeted libraries and the identification of scaffolds with high kinase inhibitory potential.

Probing drug action using in vitro pharmacological profiles.

The history of structure-activity relationships in drug design represents a long search for appropriate descriptors of broad biological action at the molecular level. In this context, recent work showing that in vitro pharmacological profiles can be used as exquisite descriptors of the broad biological effects of compounds represents an important breakthrough. Generalization of the methodology could have important implications for drug discovery and development. It might also provide a novel and insightful way to study systems biology.

Multidimensional fitting for multivariate data analysis.

Large multidimensional data matrices are frequent in biology. However, statistical methods often have difficulties dealing with such matrices because they contain very complex data sets. Consequently variable selection and dimensionality reduction methods are often used to reduce matrix complexity, although at the expense of information conservation. A new method derived from multidimensional scaling (MDS) is presented for the case where two matrices are available to describe the same population. The presented method transforms one of the matrices, called the target matrix, with some constraints to make it fit with the second matrix, referred to as the reference matrix. The fitting to the reference matrix is performed on the distances computed for the two matrices, and the transformation depends on the problem at hand. A special feature of this method is that a variable can be only partially modified. The method is applied on the exclusive-or (XOR) problem and then on a biological application with large-scale gene expression data.

Development and validation of a pharmacophore-based QSAR model for the prediction of CNS activity.

A QSAR model aimed at predicting central nervous system (CNS) activity was developed based on the structure-activity relationships of compounds from an in-house database of "drug-like" molecules. These compounds were initially identified as "CNS-active" or "CNS-inactive", and pharmacophore 3D descriptors were calculated from multiple conformations for each structure. A linear discriminant analysis (LDA) was performed on this structure-activity matrix, and this QSAR model was able to correctly classify approximately 80 % in both a learning set and a validation set. For validation purposes, the LDA model was applied to compounds for which the blood-brain barrier (BBB) penetration was known, and all of them were correctly predicted. The model was also applied to 960 other in-house compounds for which in vitro binding tests were performed on 20 receptors known to be present at the CNS level, and a high correlation was observed between compounds predicted as CNS-active and experimental hits. Finally, the model was also applied to a set of 700 structures with known CNS activity or inactivity randomly chosen from public sources, and more than 70 % of the compounds were correctly classified, including novel CNS chemotypes. These results demonstrate the applicability of the model to novel chemical structures and its usefulness for designing original CNS-focused compound libraries.

Fuzzy tricentric pharmacophore fingerprints. 1. Topological fuzzy pharmacophore triplets and adapted molecular similarity scoring schemes.

This paper introduces a novel molecular description--topological (2D) fuzzy pharmacophore triplets, 2D-FPT--using the number of interposed bonds as the measure of separation between the atoms representing pharmacophore types (hydrophobic, aromatic, hydrogen-bond donor and acceptor, cation, and anion). 2D-FPT features three key improvements with respect to the state-of-the-art pharmacophore fingerprints: (1) The first key novelty is fuzzy mapping of molecular triplets onto the basis set of pharmacophore triplets: unlike in the binary scheme where an atom triplet is set to highlight the bit of a single, best-matching basis triplet, the herein-defined fuzzy approach allows for gradual mapping of each atom triplet onto several related basis triplets, thus minimizing binary classification artifacts. (2) The second innovation is proteolytic equilibrium dependence, by explicitly considering all of the conjugated acids and bases (microspecies). 2D-FPTs are concentration-weighted (as predicted at pH=7.4) averages of microspecies fingerprints. Therefore, small structural modifications, not affecting the overall pharmacophore pattern (in the sense of classical rule-based assignment), but nevertheless triggering a pKa shift, will have a major impact on 2D-FPT. Pairs of almost identical compounds with significantly differing activities ("activity cliffs" in classical descriptor spaces) were in many cases predictable by 2D-FPT. (3) The third innovation is a new similarity scoring formula, acknowledging that the simultaneous absence of a triplet in two molecules is a less-constraining indicator of similarity than its simultaneous presence. It displays excellent neighborhood behavior, outperforming 2D or 3D two-point pharmacophore descriptors or chemical fingerprints. The 2D-FPT calculator was developed using the chemoinformatics toolkit of ChemAxon (www.chemaxon.com).

QSAR modeling of in vitro inhibition of cytochrome P450 3A4.

We report the QSAR modeling of cytochrome P450 3A4 (CYP3A4) enzyme inhibition using four large data sets of in vitro data. These data sets consist of marketed drugs and drug-like compounds all tested in four assays measuring the inhibition of the metabolism of four different substrates by the CYP3A4 enzyme. The four probe substrates are benzyloxycoumarin, testosterone, benzyloxyresorufin, and midazolam. We first show that using state-of-the-art QSAR modeling approaches applied to only one of these four data sets does not lead to predictive models that would be useful for in silico filtering of chemical libraries. We then present the development and the testing of a multiple pharmacophore hypothesis (MPH) that is formulated as a conceptual extension of the traditional QSAR approach to modeling the promiscuous binding of a large variety of drugs to CYP3A4. In the simplest form, the MPH approach takes advantage of the multiple substrate data sets and identifies the binding of test compounds as either proximal or distal relative to that of a given substrate. Application of the approach to the in silico filtering of test compounds for potential inhibitors of CYP3A4 is also presented. In addition to an improvement in the QSAR modeling for the inhibition of CYP3A4, the results from this modeling approach provide structural insights into the drug-enzyme interactions. The existence of multiple inhibition data sets in the BioPrint database motivates the original development of the concept of a multiple pharmacophore hypothesis and provides a unique opportunity for formulating alternative strategies of QSAR modeling of the inhibition of the in vitro metabolism of CYP3A4.

Structure activity relationship by NMR and by computer: a comparative study.

There has recently been considerable interest in using NMR spectroscopy to identify ligand binding sites of macromolecules. In particular, a modular approach has been put forward by Fesik et al. (Shuker, S. B.; Hajduk, P. J.; Meadows, R. P.; Fesik, S. W. Science 1996, 274, 1531-1534) in which small ligands that bind to a particular target are identified in a first round of screening and subsequently linked together to form ligands of higher affinity. Similar strategies have also been proposed for in silico drug design, where the binding sites of small chemical groups are identified, and complete ligands are subsequently assembled from different groups that have favorable interactions with the macromolecular target. In this paper, we compare experimental and computational results on a selected target (FKBP12). The binding sites of three small ligands ((2S)1-acetylprolinemethylester, 1-formylpiperidine, 1-piperidinecarboxamide) in FKBP12 were identified independently by NMR and by computational methods. The subsequent comparison of the experimental and computational data showed that the computational method identified and ranked favorably ligand positions that satisfy the experimental NOE constraints.