Automated Identification and Classification of Stereochemistry: Chirality and Double Bond Stereoisomerism

TL;DR

This paper presents a computer program and web tool for automatically analyzing and classifying stereochemistry, specifically chirality and double bond stereoisomerism, from molecular graphs to aid in chemical structure comparison.

Contribution

It introduces a novel software that automatically detects and classifies stereoisomerism in molecules based on their graph representations, focusing on explicit stereochemistry around carbon atoms.

Findings

Successfully classifies stereoisomerism types in molecular graphs

Provides a web tool for automatic stereochemistry analysis

Addresses a gap in software capabilities for stereoisomer detection

Abstract

Stereoisomers have the same molecular formula and the same atom connectivity and their existence can be related to the presence of different three-dimensional arrangements. Stereoisomerism is of great importance in many different fields since the molecular properties and biological effects of the stereoisomers are often significantly different. Most drugs for example, are often composed of a single stereoisomer of a compound, and while one of them may have therapeutic effects on the body, another may be toxic. A challenging task is the automatic detection of stereoisomers using line input specifications such as SMILES or InChI since it requires information about group theory (to distinguish stereoisomers using mathematical information about its symmetry), topology and geometry of the molecule. There are several software packages that include modules to handle stereochemistry, especially the ones to name a chemical structure and/or view, edit and generate chemical structure diagrams. However, there is a lack of software capable of automatically analyzing a molecule represented as a graph and generate a classification of the type of isomerism present in a given atom or bond. Considering the importance of stereoisomerism when comparing chemical structures, this report describes a computer program for analyzing and processing steric information contained in a chemical structure represented as a molecular graph and providing as output a binary classification of the isomer type based on the recommended conventions. Due to the complexity of the underlying issue, specification of stereochemical information is currently limited to explicit stereochemistry and to the two most common types of stereochemistry caused by asymmetry around carbon atoms: chiral atom and double bond. A Webtool to automatically identify and classify stereochemistry is available at http://nams.lasige.di.fc.ul.pt/tools.php