Peter C. Jurs

Prediction of human intestinal absorption of drug compounds from molecular structure.

The absorption of a drug compound through the human intestinal cell lining is an important property for potential drug candidates. Measuring this property, however, can be costly and time-consuming. The use of quantitative structure−property relationships (QSPRs) to estimate percent human intestinal absorption (%HIA) is an attractive alternative to experimental measurements. A data set of 86 drug and drug-like compounds with measured values of %HIA taken from the literature was used to develop and test a QSPR model. The compounds were encoded with calculated molecular structure descriptors. A nonlinear computational neural network model was developed by using the genetic algorithm with a neural network fitness evaluator. The calculated %HIA (cHIA) model performs well, with root-mean-square (rms) errors of 9.4%HIA units for the training set, 19.7%HIA units for the cross-validation (CV) set, and 16.0%HIA units for the external prediction set.

Descriptions of molecular shape applied in studies of structure/activity and structure/property relationships

Abstract Three-dimensional molecular shape is often an important determinant of the physicochemical properties or biological activities of chemical compounds. A set of six descriptors that code the shapes of molecules by calculating the areas of three orthogonal projections is developed. The descriptors are applied to a series of tests of their applicability on diverse problems in studies of structure/activity and structure/property relationships. Correlation of these descriptors with the physicochemical properties of several sets of compounds, and with biological activity values for several sets of compounds, shows that the indexes contain much information about molecular shape.

Prediction of Aqueous Solubility of Organic Compounds from Molecular Structure

Multiple linear regression (MLR) and computational neural networks (CNN) are utilized to develop mathematical models to relate the structures of a diverse set of 332 organic compounds to their aqueous solubilities. Topological, geometric, and electronic descriptors are used to numerically represent structural features of the data set compounds. Genetic algorithm and simulated annealing routines, in conjunction with MLR and CNN, are used to select subsets of descriptors that accurately relate to aqueous solubility. Nonlinear models with nine calculated structural descriptors are developed that have a training set root-mean-square error of 0.394 log units for compounds which span a −log(molarity) range from −2 to +12 log units.

Prediction of glass transition temperatures from monomer and repeat unit structure using computational neural networks

Quantitative structure-property relationships (QSPR) are developed to correlate glass transition temperatures and chemical structure. Both monomer and repeat unit structures are used to build several QSPR models for Parts 1 and 2 of this study, respectively. Models are developed using numerical descriptors, which encode important information about chemical structure (topological, electronic, and geometric). Multiple linear regression analysis (MLRA) and computational neural networks (CNNs) are used to generate the models after descriptor generation. Optimization routines (simulated annealing and genetic algorithm) are utilized to find information-rich subsets of descriptors for prediction. A 10-descriptor CNN model was found to be optimal in predicting T(g) values using the monomer structure (Part 1) for 165 polymers. A committee of 10 CNNs produced a training set rms error of 10.1K (r2 = 0.98) and a prediction set rms error of 21.7 K (r2 = 0.92). An 11-descriptor CNN model was developed for 251 polymers using the repeat unit structure (Part 2). A committee of CNNs produced a training set rms error of 21.1K (r2 = 0.96) and a prediction set rms error of 21.9 K (r2 = 0.96).

Computer-Enhanced Analytical Spectroscopy

This current volume covers research advances in nuclear magnetic resonance, mass spectrometry, and optical spectroscopy with emphasis on computer-assisted interpretation methodologies.

QSAR and k-nearest neighbor classification analysis of selective cyclooxygenase-2 inhibitors using topologically-based numerical descriptors.

Experimental IC(50) data for 314 selective cyclooxygenase-2 (COX-2) inhibitors are used to develop quantitation and classification models as a potential screening mechanism for larger libraries of target compounds. Experimental log(IC(50)) values ranged from 0.23 to > or = 5.00. Numerical descriptors encoding solely topological information are calculated for all structures and are used as inputs for linear regression, computational neural network, and classification analysis routines. Evolutionary optimization algorithms are then used to search the descriptor space for information-rich subsets which minimize the rms error of a diverse training set of compounds. An eight-descriptor model was identified as a robust predictor of experimental log(IC(50)) values, producing a root-mean-square error of 0.625 log units for an external prediction set of inhibitors which took no part in model development. A k-nearest neighbor classification study of the data set discriminating between active and inactive members produced a nine-descriptor model able to accurately classify 83.3% of the prediction set compounds correctly.

Interpreting computational neural network QSAR models : A measure of descriptor importance

We present a method to measure the relative importance of the descriptors present in a QSAR model developed with a computational neural network (CNN). The approach is based on a sensitivity analysis of the descriptors. We tested the method on three published data sets for which linear and CNN models were previously built. The original work reported interpretations for the linear models, and we compare the results of the new method to the importance of descriptors in the linear models as described by a PLS technique. The results indicate that the proposed method is able to rank descriptors such that important descriptors in the CNN model correspond to the important descriptors in the linear model.

Prediction of boiling points and critical temperatures of industrially important organic compounds from molecular structure

Numeric representations of molecular structure are used to predict the normal boiling points and critical temperatures for compounds drawn from the Design Institute for Physical Property Data (DIPPR) database. Multiple linear regression analysis and computational neural networks (i.e., using back-propagation and quasi-Newton training) are employed to develop models which can accurately predict the boiling points of 298 organic compounds. This approach is assessed by comparing its results against results obtained using the Joback group contribution approach. Finally, the same methodology is used to develop two separate critical temperature models, one based on the methods of corresponding states and the second based on structurally derived parameters alone.

Prediction of gas chromatographic retention indices of alkylbenzenes

Abstract The retention indices (RIs) of a set of alkylbenzenes on a polar gas chromatographic column are predicted directly from their molecular structures. Numerical descriptors are calculated based on the structure of a group of 150 alkylbenzenes. The descriptors are of three types: topological, geometric, and electronic. Statistical methods are employed to find an informative subset of these descriptors that can accurately predict the gas chromatographic RIs. The Automated Data Analysis and Pattern Recognition Toolkit (ADAPT) software system is used to construct a large pool of structurally derived numerical descriptors which are used to build quantitative structure-retention relationships (QSRRs). Multiple linear regression analysis and computational neural networks are used to map the descriptors to the RIs.

ADAPT: A Computer System for Automated Data Analysis Using Pattern Recognition Techniques

Listed in Table VI is a sampling of the index entries generated from coded phrases via the articulation algorithm during the course of the testing. The index entry ratings assigned by the docuiment analysts are also shown. The purpose of this listing is only to illustrate the quality of text modifications generated by the algorithm. The headings in this listing are the as-dictated headings which have not been subjected to the CAS heading terminology control systems. Therefore, in some cases, the headings shown are not the headings which would appear in the CAS volume indexes.