In-silico 3D molecular editing through physics-informed and preference-aligned generative foundation models
Abstract Generating molecular structures towards desired properties is a critical task in computer-aided drug and material design. As special 3D entities, molecules inherit non-trivial physical complexity, and many intrinsic properties may not be learnable through pure data-driven approaches, hinder...
Saved in:
| Main Authors: | , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Nature Portfolio
2025-07-01
|
| Series: | Nature Communications |
| Online Access: | https://doi.org/10.1038/s41467-025-61323-x |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1849402141471408128 |
|---|---|
| author | Xiaohan Lin Yijie Xia Yanheng Li Yu-Peng Huang Shuo Liu Jun Zhang Yi Qin Gao |
| author_facet | Xiaohan Lin Yijie Xia Yanheng Li Yu-Peng Huang Shuo Liu Jun Zhang Yi Qin Gao |
| author_sort | Xiaohan Lin |
| collection | DOAJ |
| description | Abstract Generating molecular structures towards desired properties is a critical task in computer-aided drug and material design. As special 3D entities, molecules inherit non-trivial physical complexity, and many intrinsic properties may not be learnable through pure data-driven approaches, hindering the transaction of powerful generative artificial intelligence (GenAI) to this field. To avoid existing molecular GenAI’s heavy reliance on domain-specific models and priors, in this research, we derive theoretical guidelines to bridge the methodological gap between GenAI for images and molecules, allowing pre-training of foundation models for 3D molecular generation. Difficulties due to symmetry, stability and entropy, which are critical for molecules, are overcome through a simple and model-agnostic training protocol. Moreover, we apply physics-informed strategies to force MolEdit, a pre-trained multimodal molecular GenAI, to obey physics laws and align with contextual preferences, and thus suppress undesired model hallucinations. MolEdit can generate valid molecules with comprehensive symmetry, strikes a better balance between configuration stability and conformer diversity, and supports complicated 3D scaffolds which frustrate other methods. Furthermore, MolEdit is applicable for zero-shot lead optimization and linker design following contextual and geometrical specifications. Collectively, as a foundation model, MolEdit offers flexibility and developability for AI-aided editing and manipulation of molecules serving various purposes. |
| format | Article |
| id | doaj-art-bad0cf75407a4150a9ee82bd07331f02 |
| institution | Kabale University |
| issn | 2041-1723 |
| language | English |
| publishDate | 2025-07-01 |
| publisher | Nature Portfolio |
| record_format | Article |
| series | Nature Communications |
| spelling | doaj-art-bad0cf75407a4150a9ee82bd07331f022025-08-20T03:37:37ZengNature PortfolioNature Communications2041-17232025-07-0116111510.1038/s41467-025-61323-xIn-silico 3D molecular editing through physics-informed and preference-aligned generative foundation modelsXiaohan Lin0Yijie Xia1Yanheng Li2Yu-Peng Huang3Shuo Liu4Jun Zhang5Yi Qin Gao6New Cornerstone Science Laboratory, Beijing National Laboratory for Molecular Sciences, College of Chemistry and Molecular Engineering, Peking UniversityNew Cornerstone Science Laboratory, Beijing National Laboratory for Molecular Sciences, College of Chemistry and Molecular Engineering, Peking UniversityNew Cornerstone Science Laboratory, Beijing National Laboratory for Molecular Sciences, College of Chemistry and Molecular Engineering, Peking UniversityNew Cornerstone Science Laboratory, Beijing National Laboratory for Molecular Sciences, College of Chemistry and Molecular Engineering, Peking UniversitySchool of Pharmacy, Lanzhou UniversityChangping LaboratoryNew Cornerstone Science Laboratory, Beijing National Laboratory for Molecular Sciences, College of Chemistry and Molecular Engineering, Peking UniversityAbstract Generating molecular structures towards desired properties is a critical task in computer-aided drug and material design. As special 3D entities, molecules inherit non-trivial physical complexity, and many intrinsic properties may not be learnable through pure data-driven approaches, hindering the transaction of powerful generative artificial intelligence (GenAI) to this field. To avoid existing molecular GenAI’s heavy reliance on domain-specific models and priors, in this research, we derive theoretical guidelines to bridge the methodological gap between GenAI for images and molecules, allowing pre-training of foundation models for 3D molecular generation. Difficulties due to symmetry, stability and entropy, which are critical for molecules, are overcome through a simple and model-agnostic training protocol. Moreover, we apply physics-informed strategies to force MolEdit, a pre-trained multimodal molecular GenAI, to obey physics laws and align with contextual preferences, and thus suppress undesired model hallucinations. MolEdit can generate valid molecules with comprehensive symmetry, strikes a better balance between configuration stability and conformer diversity, and supports complicated 3D scaffolds which frustrate other methods. Furthermore, MolEdit is applicable for zero-shot lead optimization and linker design following contextual and geometrical specifications. Collectively, as a foundation model, MolEdit offers flexibility and developability for AI-aided editing and manipulation of molecules serving various purposes.https://doi.org/10.1038/s41467-025-61323-x |
| spellingShingle | Xiaohan Lin Yijie Xia Yanheng Li Yu-Peng Huang Shuo Liu Jun Zhang Yi Qin Gao In-silico 3D molecular editing through physics-informed and preference-aligned generative foundation models Nature Communications |
| title | In-silico 3D molecular editing through physics-informed and preference-aligned generative foundation models |
| title_full | In-silico 3D molecular editing through physics-informed and preference-aligned generative foundation models |
| title_fullStr | In-silico 3D molecular editing through physics-informed and preference-aligned generative foundation models |
| title_full_unstemmed | In-silico 3D molecular editing through physics-informed and preference-aligned generative foundation models |
| title_short | In-silico 3D molecular editing through physics-informed and preference-aligned generative foundation models |
| title_sort | in silico 3d molecular editing through physics informed and preference aligned generative foundation models |
| url | https://doi.org/10.1038/s41467-025-61323-x |
| work_keys_str_mv | AT xiaohanlin insilico3dmoleculareditingthroughphysicsinformedandpreferencealignedgenerativefoundationmodels AT yijiexia insilico3dmoleculareditingthroughphysicsinformedandpreferencealignedgenerativefoundationmodels AT yanhengli insilico3dmoleculareditingthroughphysicsinformedandpreferencealignedgenerativefoundationmodels AT yupenghuang insilico3dmoleculareditingthroughphysicsinformedandpreferencealignedgenerativefoundationmodels AT shuoliu insilico3dmoleculareditingthroughphysicsinformedandpreferencealignedgenerativefoundationmodels AT junzhang insilico3dmoleculareditingthroughphysicsinformedandpreferencealignedgenerativefoundationmodels AT yiqingao insilico3dmoleculareditingthroughphysicsinformedandpreferencealignedgenerativefoundationmodels |