The Coding of the Three-Dimensional Structure of Molecules by Molecular Transforms and Its Application to Structure-Spectra Correlations and Studies of Biological Activity

TLDR

The study develops a molecular transform that encodes 3‑D molecular structures into a fixed set of values. The method flexibly incorporates atomic properties and is applied in counterpropagation neural networks to classify dopamine agonists and steroid binding affinities. The 3D‑MoRSE representation preserves key structural features, differentiates aromatic and aliphatic compounds, and shows potential for infrared spectrum simulation.

Abstract

A molecular transform, derived from an equation used in electron diffraction studies, is developed that allows the representation of the three-dimensional structure of a molecule by a fixed number of values. Various atomic properties can be taken into account giving high flexibility to this representation of a molecule. This 3D-MoRSE (Molecule Representation of Structures based on Electron diffraction) code retains important structural features such as the mass (see ref 35) and the amount of branching as evidenced by an investigation of monosubstituted benzene derivatives. Furthermore, this molecular representation was able to distinguish between benzene, cyclohexane, and naphthalene derivatives in a dataset of great structural variety. This molecular representation was used in counterpropagation neural networks to distinguish between dopamine D1 and D2 agonists and to group 31 steroids binding to the corticosteroid binding globulin receptor into compounds of high, medium, and low activity. Great promise is given to this representation of molecular structures for the simulation of infrared spectra as revealed by an investigation of monosubstituted benzene derivatives.