-
[show abstract]
[hide abstract]
ABSTRACT: The 1:1 equilibrium constants, K, for the association of hydrogen bond bases and hydrogen bond acids have been determined by using octan-1-ol solvent at 298 K for 30 acid-base combinations. The values of K are much smaller than those found for aprotic, rather nonpolar solvents. It is shown that the log K values can satisfactorily be correlated against α(H)(2)·β(H)(2), where α(H)(2) and β(H)(2) are the 1:1 hydrogen bond acidities and basicities of solutes. The slope of the plot, 2.938, is much smaller than those for log K values in the nonpolar organic solvents previously studied. An analysis of literature data on 1:1 hydrogen bonding in water yields a negative slope for a plot of log K against α(H)(2)·β(H)(2), thus showing how the use of very strong hydrogen bond acids and bases does not lead to larger values of log K for 1:1 hydrogen bonding in water. It is suggested that for simple 1:1 association between monofunctional solutes in water, log K cannot be larger than about -0.1 log units. Descriptors have been obtained for the complex between 2,2,2-trifluoroethanol and propanone, and used to analyze solvent effects on the two reactants, the complex, and the complexation constant.
The Journal of Organic Chemistry 10/2010; · 4.45 Impact Factor
-
[show abstract]
[hide abstract]
ABSTRACT: There are no previous references to the direct use of GLC data in the correlation of biological processes, but we show that GLC retention data can be used in the correlation of several such processes involving gaseous solutes. There are a number of reports of RP-HPLC and MEKC data being used in the correlation of biological processes, but they are mostly restricted as to the number and type of solute studied. We show that if chromatographic data are used to obtain solvation descriptors for solutes, and if these descriptors are then used in the correlation of biological processes, that this indirect connection is a much more powerful and generally applicable method than is the direct connection between chromatographic data and biological data.
Journal of Chromatography B: Biomedical Sciences and Applications.
-
[show abstract]
[hide abstract]
ABSTRACT: In this article, we present an automatic model generation process for building QSAR models using Gaussian Processes, a powerful machine learning modeling method. We describe the stages of the process that ensure models are built and validated within a rigorous framework: descriptor calculation, splitting data into training, validation and test sets, descriptor filtering, application of modeling techniques and selection of the best model. We apply this automatic process to data sets of blood-brain barrier penetration and aqueous solubility and compare the resulting automatically generated models with 'manually' built models using external test sets. The results demonstrate the effectiveness of the automatic model generation process for two types of data sets commonly encountered in building ADME QSAR models, a small set of in vivo data and a large set of physico-chemical data.
Journal of Computer-Aided Molecular Design 22(6-7):431-40. · 3.39 Impact Factor
-
[show abstract]
[hide abstract]
ABSTRACT: In this article, we discuss the application of the Gaussian Process method for the prediction of absorption, distribution, metabolism, and excretion (ADME) properties. On the basis of a Bayesian probabilistic approach, the method is widely used in the field of machine learning but has rarely been applied in quantitative structure-activity relationship and ADME modeling. The method is suitable for modeling nonlinear relationships, does not require subjective determination of the model parameters, works for a large number of descriptors, and is inherently resistant to overtraining. The performance of Gaussian Processes compares well with and often exceeds that of artificial neural networks. Due to these features, the Gaussian Processes technique is eminently suitable for automatic model generation-one of the demands of modern drug discovery. Here, we describe the basic concept of the method in the context of regression problems and illustrate its application to the modeling of several ADME properties: blood-brain barrier, hERG inhibition, and aqueous solubility at pH 7.4. We also compare Gaussian Processes with other modeling techniques.
Journal of Chemical Information and Modeling 47(5):1847-57. · 4.68 Impact Factor
-
[show abstract]
[hide abstract]
ABSTRACT: This paper describes the development of an aqueous solubility model based on solubility data from the Syracuse database, calculated octanol-water partition coefficient, and 51 2D molecular descriptors. Two different statistical packages, SIMCA and Cubist, were used and the results were compared. The Cubist model, which comprises a collection of rules, each of which has an associated Multiple Linear Regression model (MLR), gave better overall results on a test set of 640 compounds with an overall squared correlation coefficient of 0.74 and an absolute average error of 0.68 log units. Both training and independent test sets had similar distributions of structures in terms of the different functionalities present-60% neutral, 14% acidic, 8% phenolic, 11% monobasic, 4% polybasic, and 3% zwitterionic molecules. Sets were designed by random selection, with 2688 (81%) and 640 (19%) molecules, respectively, forming the training and the test sets.
Journal of Chemical Information and Computer Sciences 43(3):837-41.
-
[show abstract]
[hide abstract]
ABSTRACT: We measured psychometric (i.e. concentration-response) functions for the detection of odor, nasal pungency and eye irritation from butyl acetate and toluene. Olfactory detection was measured in subjects with normal olfaction (i.e. normosmics) for whom nasal trigeminal detection does not interfere because it requires much higher concentrations. Nasal trigeminal detection, called nasal pungency, was measured only in subjects lacking olfaction (i.e. anosmics) in order to avoid odor interference. Ocular trigeminal detection, called eye irritation, was measured in both groups. The method employed entailed a two-alternative, forced-choice procedure with presentation of increasing concentrations. The outcome showed, for both chemicals, similar ocular trigeminal chemosensitivity in normosmics and anosmics and similar overall ocular and nasal trigeminal chemosensitivity. Olfactory sensitivity was much higher than both forms of trigeminal sensitivity by concentration differences of six and four orders of magnitude for butyl acetate and toluene, respectively. Detectability plots (i.e. detection performance vs log concentration) for the three sensory endpoints followed an S-shaped function with a middle range section that showed a robust linear fit (r > 0.94) on graphs of z-score vs log concentration. These detectability functions allow the calculation of olfactory and trigeminal thresholds at various levels of performance. At a point half-way between random and perfect detection, trigeminal and olfactory threshold concentrations were, respectively, 0.67 (+/-0.32) and 2.28 (+/-1.77) log units lower than those measured by us in the past for the same chemicals using an analogous procedure but under just one, fixed, level of performance. The available data suggest that, although considerably laborious, detectability functions provide chemosensory thresholds of closer relevance to environmentally realistic conditions (e.g. whole-body exposures).
Journal of Applied Toxicology 22(1):25-30. · 2.48 Impact Factor
