Stereoisomers Are Not Machine Learning's Best Friends.

Tahıl, G.; Delorme, F.; Le Berre, D.; Monflier, Eric; Sayede, Adlane; Tilloy, Sebastien

Document type :

Article dans une revue scientifique: Article original

DOI :

10.1021/acs.jcim.4c00318

PMID :

38949069

Permalink :

http://hdl.handle.net/20.500.12210/116130

Title :

Stereoisomers Are Not Machine Learning's Best Friends.

Author(s) :

Tahıl, G. [Auteur]
Centre de Recherche en Informatique de Lens [CRIL]
Unité de Catalyse et Chimie du Solide - UMR 8181 [UCCS]
Delorme, F. [Auteur]
Centre de Recherche en Informatique de Lens [CRIL]
Le Berre, D. [Auteur]
Centre de Recherche en Informatique de Lens [CRIL]
Monflier, Eric [Auteur]

Unité de Catalyse et Chimie du Solide (UCCS) - UMR 8181
Sayede, Adlane [Auteur]

Unité de Catalyse et Chimie du Solide - UMR 8181 [UCCS]
Tilloy, Sebastien [Auteur]

Unité de Catalyse et Chimie du Solide - UMR 8181 [UCCS]

Journal title :

J Chem Inf Model

Abbreviated title :

J Chem Inf Model

Volume number :

Pages :

5451–5469

Publication date :

2024-07-05

ISSN :

1549-960X

English keyword(s) :

Algorithms
Molecular modeling
Molecular structure
Molecules
Stereochemistry

HAL domain(s) :

Chimie/Catalyse

English abstract : [en]

This study addresses the challenge of accurately identifying stereoisomers in cheminformatics, which originates from our objective to apply machine learning to predict the association constant between cyclodextrin and a ...
Show more >This study addresses the challenge of accurately identifying stereoisomers in cheminformatics, which originates from our objective to apply machine learning to predict the association constant between cyclodextrin and a guest. Identifying stereoisomers is indeed crucial for machine learning applications. Current tools offer various molecular descriptors, including their textual representation as Isomeric SMILES that can distinguish stereoisomers. However, such representation is text-based and does not have a fixed size, so a conversion is needed to make it usable to machine learning approaches. Word embedding techniques can be used to solve this problem. Mol2vec, a word embedding approach for molecules, offers such a conversion. Unfortunately, it cannot distinguish between stereoisomers due to its inability to capture the spatial configuration of molecular structures. This study proposes several approaches that use word embedding techniques to handle molecular discrimination using stereochemical information on molecules or considering Isomeric SMILES notation as a text in Natural Language Processing. Our aim is to generate a distinct vector for each unique molecule, correctly identifying stereoisomer information in cheminformatics. The proposed approaches are then compared to our original machine learning task: predicting the association constant between cyclodextrin and a guest molecule.Show less >

Language :

Anglais

Audience :

Internationale

Popular science :

Non

Administrative institution(s) :

Université de Lille
CNRS
Centrale Lille
ENSCL
Univ. Artois

Collections :

Unité de Catalyse et Chimie du Solide (UCCS) - UMR 8181

Submission date :

2024-07-23T21:01:50Z
2024-08-23T09:20:03Z

Version	Link	Modification date
2	20.500.12210/116130*	2024-08-23T09:13:17Z
1	20.500.12210/116130.1	2024-07-23T21:01:50Z

Stereoisomers Are Not Machine Learning's ...

BibTeX

CSV

Excel

RIS

Version history

Stereoisomers Are Not Machine Learning's ... BibTeX CSV Excel RIS

Version history

Stereoisomers Are Not Machine Learning's ...

BibTeX

CSV

Excel

RIS