Descriptors, Physical Properties, and Drug-Likeness

Abstract

We have investigated techniques for distinguishing between drugs and nondrugs using a set of molecular descriptors derived from semiempirical molecular orbital (AM1) calculations. The "drug" data set of 2105 compounds was derived from the World Drug Index (WDI) using a procedure designed to select real drugs. The "nondrug" data set was the Maybridge database. We have first investigated the dimensionality of physical properties space based on a set of 26 descriptors that we have used successfully to build absorption, distribution, metabolism, and excretion-related quantitative structure-property relationship models. We discuss the general nature of the descriptors for physical property space and the ability of these descriptors to distinguish between drugs and nondrugs. The third most significant principal component of this set of descriptors serves as a useful numerical index of drug-likeness, but no others are able to distinguish between drugs and nondrugs. We have therefore extended our set of descriptors to a total of 66 and have used recursive partitioning to identify the descriptors that can distinguish between drugs and nondrugs. This procedure pointed to two of the descriptors that play an important role in the principal component found above and one more from the set of 40 extra descriptors. These three descriptors were then used to train a Kohonen artificial neural net for the entire Maybridge data set. Projecting the drug database onto the map obtained resulted in a clear distinction not only between drugs and nondrugs but also, for instance, between hormones and other drugs. Projection of 42 131 compounds from the WDI onto the Kohonen map also revealed pronounced clustering in the regions of the map assigned as druglike.

Local molecular properties and their use in predicting reactivity

“…These properties are fundamentally different to the MEP because they are not physically observable, but nevertheless provide information about regio-or stereoselectivity of reactions, the relative reactivity of different substrates etc. Murray and Politzer [6][7][8][9][10] have used the MEP projected onto the isodensity molecular surface to derive descriptors for predicting the physical properties of molecules and we [11][12][13][14][15][16][17][18] have used such descriptors calculated with semiempirical molecular orbital (MO) theory extensively in quantitative structure-property relationships (QSPRs). These descriptors however, only treat electrostatic, not donor/acceptor interactions and therefore do not provide a complete description of the surface properties of molecules.…”

QSAR and QSPR based solely on surface properties?

“…This situation is not always easy to detect because for some key properties, such as aqueous solubility, the published experimental data is itself very local and moreover does not include important classes of compounds, for instance drugs. This is emphasized by the physical property map [25,26] shown in Figure 2. The colorcoded points mark the positions of compounds from the Aquasol database [27] whose aqueous solubility is known at 298K.…”

Clustering Genes, Tissues, Cells and Bioactive Chemicals by Sphere SOM

“…Especially, ligand-based method is suitable for selecting drug candidates from enormous compounds library because it simply requires computational resources, which are less expensive. Although SOM has partly been used as a tool of ligand-based methods to classify compounds by their properties in chemoinformatics, spherical SOM has not been used in chemoinformatics to the best of our knowledge (Brüstle et al, 2002;Schneider & Nettekoven, 2003;Wang et al, 2005;Kaiser et al, 2007;Renner et al, 2007;Li & Gramatica, 2010). We propose here the extended application of sSOM to classify bioactive compounds by their MOA together with their structural information.…”