Novel methods for the prediction of logP,pK(a), and logD

Novel methods for predicting logP, pK(a), and logD values have been developed using data sets (592 molecules for logP and 1029 for pK(a)) containing a wide range of molecular structures. An equation with three molecular properties (polarizability and partial atomic charges on nitrogen and oxygen) correlates highly with logP (r2 = 0.89). The pK(a)s are estimated for both acids and bases using a novel tree structured fingerprint describing the ionizing centers. The new models have been compared with existing models and also experimental measurements on test sets of common organic compounds and pharmaceutical molecules.     

Fragment-based quantitative structure-activity relationship (FB-QSAR) for fragment-based drug design

In cooperation with the fragment-based design a new drug design method, the so-called "fragment-based quantitative structure-activity relationship" (FB-QSAR) is proposed. The essence of the new method is that the molecular framework in a family of drug candidates are divided into several fragments according to their substitutes being investigated. The bioactivities of molecules are correlated with the physicochemical properties of the molecular fragments through two sets of coefficients in the linear free energy equations. One coefficient set is for the physicochemical properties and the other for the weight factors of the molecular fragments. Meanwhile, an iterative double least square (IDLS) technique is developed to solve the two sets of coefficients in a training data set alternately and iteratively. The IDLS technique is a feedback procedure with machine learning ability. The standard Two-dimensional quantitative structure-activity relationship (2D-QSAR) is a special case, in the FB-QSAR, when the whole molecule is treated as one entity. The FB-QSAR approach can remarkably enhance the predictive power and provide more structural insights into rational drug design. As an example, the FB-QSAR is applied to build a predictive model of neuraminidase inhibitors for drug development against H5N1 influenza virus.     

Analysis and prediction of combinatorial chemistry synthesis and screening data

The goal of combinatorial chemistry is to simultaneously synthesize sets of compounds possessing properties that are then distinguished through screening. As the size of a compound set increases, data analysis becomes more challenging. Analysis of Variance (ANOVA) is an accepted statistical method that offers a straightforward solution to this problem. Two steps encountered by combinatorial scientists appear well suited to ANOVA: the prediction of synthetic outcomes (purity and yield) of set members and the analysis of screening data to identify combinations of reagent inputs that result in molecules with a desired property. To illustrate, a subset of a combinatorial array, referred to as a reaction rehearsal set, is evaluated to create a model predictive of the individual synthetic outcomes of the full matrix. In a second exercise, the biochemical screening data obtained from a combinatorial library is analyzed to identify reagent interactions that result in molecules possessing the sought activity.     

One- to four-dimensional kernels for virtual screening and the prediction of physical, chemical, and biological properties

Many chemoinformatics applications, including high-throughput virtual screening, benefit from being able to rapidly predict the physical, chemical, and biological properties of small molecules to screen large repositories and identify suitable candidates. When training sets are available, machine learning methods provide an effective alternative to ab initio methods for these predictions. Here, we leverage rich molecular representations including 1D SMILES strings, 2D graphs of bonds, and 3D coordinates to derive efficient machine learning kernels to address regression problems. We further expand the library of available spectral kernels for small molecules developed for classification problems to include 2.5D surface and 3D kernels using Delaunay tetrahedrization and other techniques from computational geometry, 3D pharmacophore kernels, and 3.5D or 4D kernels capable of taking into account multiple molecular configurations, such as conformers. The kernels are comprehensively tested using cross-validation and redundancy-reduction methods on regression problems using several available data sets to predict boiling points, melting points, aqueous solubility, octanol/water partition coefficients, and biological activity with state-of-the art results. When sufficient training data are available, 2D spectral kernels in general tend to yield the best and most robust results, better than state-of-the art. On data sets containing thousands of molecules, the kernels achieve a squared correlation coefficient of 0.91 for aqueous solubility prediction and 0.94 for octanol/water partition coefficient prediction. Averaging over conformations improves the performance of kernels based on the three-dimensional structure of molecules, especially on challenging data sets. Kernel predictors for aqueous solubility (kSOL), LogP (kLOGP), and melting point (kMELT) are available over the Web through: http://cdb.ics.uci.edu.     

Ligand-target interaction-based weighting of substructures for virtual screening

A methodology is introduced to assign energy-based scores to two-dimensional (2D) structural features based on three-dimensional (3D) ligand-target interaction information and utilize interaction-annotated features in virtual screening. Database molecules containing such fragments are assigned cumulative scores that serve as a measure of similarity to active reference compounds. The Interaction Annotated Structural Features (IASF) method is applied to mine five high-throughput screening (HTS) data sets and often identifies more hits than conventional fragment-based similarity searching or ligand-protein docking.     

ESOL: estimating aqueous solubility directly from molecular structure

This paper describes a simple method for estimating the aqueous solubility (ESOL--Estimated SOLubility) of a compound directly from its structure. The model was derived from a set of 2874 measured solubilities using linear regression against nine molecular properties. The most significant parameter was calculated logP(octanol), followed by molecular weight, proportion of heavy atoms in aromatic systems, and number of rotatable bonds. The model performed consistently well across three validation sets, predicting solubilities within a factor of 5-8 of their measured values, and was competitive with the well-established "General Solubility Equation" for medicinal/agrochemical sized molecules.     

New fragment weighting scheme for the Bayesian inference network in ligand-based virtual screening

Many of the conventional similarity methods assume that molecular fragments that do not relate to biological activity carry the same weight as the important ones. One possible approach to this problem is to use the Bayesian inference network (BIN), which models molecules and reference structures as probabilistic inference networks. The relationships between molecules and reference structures in the Bayesian network are encoded using a set of conditional probability distributions, which can be estimated by the fragment weighting function, a function of the frequencies of the fragments in the molecule or the reference structure as well as throughout the collection. The weighting function combines one or more fragment weighting schemes. In this paper, we have investigated five different weighting functions and present a new fragment weighting scheme. Later on, these functions were modified to combine the new weighting scheme. Simulated virtual screening experiments with the MDL Drug Data Report (23) and maximum unbiased validation data sets show that the use of new weighting scheme can provide significantly more effective screening when compared with the use of current weighting schemes.