Deep generative models for 3D linker design
Rational compound design remains a challenging problem for both computational methods and medicinal chemists. Computational generative methods have begun to show promising results for the design problem. However, they have not yet used the power of three-dimensional (3D) structural information. We h...
Main Authors: | , , , |
---|---|
Format: | Journal article |
Language: | English |
Published: |
American Chemical Society
2020
|
_version_ | 1797061722162331648 |
---|---|
author | Imrie, F Bradley, AR van der Schaar, M Deane, CM |
author_facet | Imrie, F Bradley, AR van der Schaar, M Deane, CM |
author_sort | Imrie, F |
collection | OXFORD |
description | Rational compound design remains a challenging problem for both computational methods and medicinal chemists. Computational generative methods have begun to show promising results for the design problem. However, they have not yet used the power of three-dimensional (3D) structural information. We have developed a novel graph-based deep generative model that combines state-of-the-art machine learning techniques with structural knowledge. Our method (“DeLinker”) takes two fragments or partial structures and designs a molecule incorporating both. The generation process is protein-context-dependent, utilizing the relative distance and orientation between the partial structures. This 3D information is vital to successful compound design, and we demonstrate its impact on the generation process and the limitations of omitting such information. In a large-scale evaluation, DeLinker designed 60% more molecules with high 3D similarity to the original molecule than a database baseline. When considering the more relevant problem of longer linkers with at least five atoms, the outperformance increased to 200%. We demonstrate the effectiveness and applicability of this approach on a diverse range of design problems: fragment linking, scaffold hopping, and proteolysis targeting chimera (PROTAC) design. As far as we are aware, this is the first molecular generative model to incorporate 3D structural information directly in the design process. The code is available at https://github.com/oxpig/DeLinker. |
first_indexed | 2024-03-06T20:35:19Z |
format | Journal article |
id | oxford-uuid:326f715e-ec18-4683-bf77-cf84338f3be0 |
institution | University of Oxford |
language | English |
last_indexed | 2024-03-06T20:35:19Z |
publishDate | 2020 |
publisher | American Chemical Society |
record_format | dspace |
spelling | oxford-uuid:326f715e-ec18-4683-bf77-cf84338f3be02022-03-26T13:14:01ZDeep generative models for 3D linker designJournal articlehttp://purl.org/coar/resource_type/c_dcae04bcuuid:326f715e-ec18-4683-bf77-cf84338f3be0EnglishSymplectic ElementsAmerican Chemical Society 2020Imrie, FBradley, ARvan der Schaar, MDeane, CMRational compound design remains a challenging problem for both computational methods and medicinal chemists. Computational generative methods have begun to show promising results for the design problem. However, they have not yet used the power of three-dimensional (3D) structural information. We have developed a novel graph-based deep generative model that combines state-of-the-art machine learning techniques with structural knowledge. Our method (“DeLinker”) takes two fragments or partial structures and designs a molecule incorporating both. The generation process is protein-context-dependent, utilizing the relative distance and orientation between the partial structures. This 3D information is vital to successful compound design, and we demonstrate its impact on the generation process and the limitations of omitting such information. In a large-scale evaluation, DeLinker designed 60% more molecules with high 3D similarity to the original molecule than a database baseline. When considering the more relevant problem of longer linkers with at least five atoms, the outperformance increased to 200%. We demonstrate the effectiveness and applicability of this approach on a diverse range of design problems: fragment linking, scaffold hopping, and proteolysis targeting chimera (PROTAC) design. As far as we are aware, this is the first molecular generative model to incorporate 3D structural information directly in the design process. The code is available at https://github.com/oxpig/DeLinker. |
spellingShingle | Imrie, F Bradley, AR van der Schaar, M Deane, CM Deep generative models for 3D linker design |
title | Deep generative models for 3D linker design |
title_full | Deep generative models for 3D linker design |
title_fullStr | Deep generative models for 3D linker design |
title_full_unstemmed | Deep generative models for 3D linker design |
title_short | Deep generative models for 3D linker design |
title_sort | deep generative models for 3d linker design |
work_keys_str_mv | AT imrief deepgenerativemodelsfor3dlinkerdesign AT bradleyar deepgenerativemodelsfor3dlinkerdesign AT vanderschaarm deepgenerativemodelsfor3dlinkerdesign AT deanecm deepgenerativemodelsfor3dlinkerdesign |