Comparisons of Molecular Structure Generation Methods Based on Fragment Assemblies and Genetic Graphs - Archive ouverte HAL Access content directly
Journal Articles Journal of Chemical Information and Modeling Year : 2021

Comparisons of Molecular Structure Generation Methods Based on Fragment Assemblies and Genetic Graphs

(1) , (1) , (1)
1
Benoit Creton
Carlos Nieto-Draghi

Abstract

The use of quantitative structure–property relationships (QSPRs) helps in predicting molecular properties for several decades, while the automatic design of new molecular structures is still emerging. The choice of algorithms to generate molecules is not obvious and is related to several factors such as the desired chemical diversity (according to an initial dataset’s content) and the level of construction (the use of atoms, fragments, pattern-based methods). In this paper, we address the problem of molecular structure generation by revisiting two approaches: fragment-based methods (FMs) and genetic-based methods (GMs). We define a set of indices to compare generation methods on a specific task. New indices inform about the explored data space (coverage), compare how the data space is explored (representativeness), and quantifies the ratio of molecules satisfying requirements (generation specificity) without the use of a database composed of real chemicals as a reference. These indices were employed to compare generations of molecules fulfilling the desired property criterion, evaluated by QSPR.
Fichier principal
Vignette du fichier
Comparisons of Molecular Structures Generation Methods Based on Fragments Assemblies and Genetic Graphs.pdf (8.04 Mo) Télécharger le fichier
Vignette du fichier
Comparisons of Molecular Structures Generation Sup mat bis.pdf (352.73 Ko) Télécharger le fichier
Vignette du fichier
Gantze-SI2.xlsx (239.1 Ko) Télécharger le fichier
Origin : Files produced by the author(s)

Dates and versions

hal-03498045 , version 1 (20-12-2021)

Identifiers

Cite

Philippe Gantzer, Benoit Creton, Carlos Nieto-Draghi. Comparisons of Molecular Structure Generation Methods Based on Fragment Assemblies and Genetic Graphs. Journal of Chemical Information and Modeling, 2021, 61 (9), pp.4245-4258. ⟨10.1021/acs.jcim.1c00803⟩. ⟨hal-03498045⟩

Collections

IFP
41 View
43 Download

Altmetric

Share

Gmail Facebook Twitter LinkedIn More