Explainable Fragment‐Based Molecular Property Attribution
“AI & Drug Discovery” mode has significantly promoted drug development and achieved excellent performance, especially with the rapid development of deep learning, making remarkable contributions to protecting human physiological health. However, due to the “black‐box” characteristic of the deep...
| Main Authors: | , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Wiley
2022-10-01
|
| Series: | Advanced Intelligent Systems |
| Subjects: | |
| Online Access: | https://doi.org/10.1002/aisy.202200104 |
| _version_ | 1798027638079488000 |
|---|---|
| author | Lingxiang Jia Zunlei Feng Haotian Zhang Jie Song Zipeng Zhong Shaolun Yao Mingli Song |
| author_facet | Lingxiang Jia Zunlei Feng Haotian Zhang Jie Song Zipeng Zhong Shaolun Yao Mingli Song |
| author_sort | Lingxiang Jia |
| collection | DOAJ |
| description | “AI & Drug Discovery” mode has significantly promoted drug development and achieved excellent performance, especially with the rapid development of deep learning, making remarkable contributions to protecting human physiological health. However, due to the “black‐box” characteristic of the deep learning model, the decision route and predicted results in different research stages assisted by deep models are usually unexplainable, limiting their application in practice and more in‐depth research of drug discovery. Focusing on the drug molecules, an explainable fragment‐based molecular property attribution technique is proposed for analyzing the influence of particular molecule fragments on properties and the relationship between the molecular properties herein. Quantitative experiments on 42 benchmark property tasks demonstrate that 325 attribution fragments, which account for 90% of the overall attribution results obtained by the proposed method, have positive relevance to the corresponding property tasks. More impressively, most of the attribution results randomly selected are consistent with the existing mechanism explanations. The discovery mentioned above provides a reference standard for assisting researchers in developing more specific and practical drug molecule studies, such as synthesizing molecule with the targeted property using a fragment obtained from the attribution method. An interactive preprint version of the article can be found at: https://www.authorea.com/doi/full/10.22541/au.165279262.29589148. |
| first_indexed | 2024-04-11T18:54:46Z |
| format | Article |
| id | doaj.art-fe7732f7bf4b498d9eeca70cacb8f419 |
| institution | Directory Open Access Journal |
| issn | 2640-4567 |
| language | English |
| last_indexed | 2024-04-11T18:54:46Z |
| publishDate | 2022-10-01 |
| publisher | Wiley |
| record_format | Article |
| series | Advanced Intelligent Systems |
| spelling | doaj.art-fe7732f7bf4b498d9eeca70cacb8f4192022-12-22T04:08:13ZengWileyAdvanced Intelligent Systems2640-45672022-10-01410n/an/a10.1002/aisy.202200104Explainable Fragment‐Based Molecular Property AttributionLingxiang Jia0Zunlei Feng1Haotian Zhang2Jie Song3Zipeng Zhong4Shaolun Yao5Mingli Song6College of Computer Science Zhejiang University Hangzhou 310027 ChinaCollege of Computer Science Zhejiang University Hangzhou 310027 ChinaCollege of Pharmaceutical Sciences Zhejiang University Hangzhou 310027 ChinaCollege of Computer Science Zhejiang University Hangzhou 310027 ChinaCollege of Computer Science Zhejiang University Hangzhou 310027 ChinaCollege of Computer Science Zhejiang University Hangzhou 310027 ChinaCollege of Computer Science Zhejiang University Hangzhou 310027 China“AI & Drug Discovery” mode has significantly promoted drug development and achieved excellent performance, especially with the rapid development of deep learning, making remarkable contributions to protecting human physiological health. However, due to the “black‐box” characteristic of the deep learning model, the decision route and predicted results in different research stages assisted by deep models are usually unexplainable, limiting their application in practice and more in‐depth research of drug discovery. Focusing on the drug molecules, an explainable fragment‐based molecular property attribution technique is proposed for analyzing the influence of particular molecule fragments on properties and the relationship between the molecular properties herein. Quantitative experiments on 42 benchmark property tasks demonstrate that 325 attribution fragments, which account for 90% of the overall attribution results obtained by the proposed method, have positive relevance to the corresponding property tasks. More impressively, most of the attribution results randomly selected are consistent with the existing mechanism explanations. The discovery mentioned above provides a reference standard for assisting researchers in developing more specific and practical drug molecule studies, such as synthesizing molecule with the targeted property using a fragment obtained from the attribution method. An interactive preprint version of the article can be found at: https://www.authorea.com/doi/full/10.22541/au.165279262.29589148.https://doi.org/10.1002/aisy.202200104drug discoveryexplainable machine learningmolecular properties |
| spellingShingle | Lingxiang Jia Zunlei Feng Haotian Zhang Jie Song Zipeng Zhong Shaolun Yao Mingli Song Explainable Fragment‐Based Molecular Property Attribution Advanced Intelligent Systems drug discovery explainable machine learning molecular properties |
| title | Explainable Fragment‐Based Molecular Property Attribution |
| title_full | Explainable Fragment‐Based Molecular Property Attribution |
| title_fullStr | Explainable Fragment‐Based Molecular Property Attribution |
| title_full_unstemmed | Explainable Fragment‐Based Molecular Property Attribution |
| title_short | Explainable Fragment‐Based Molecular Property Attribution |
| title_sort | explainable fragment based molecular property attribution |
| topic | drug discovery explainable machine learning molecular properties |
| url | https://doi.org/10.1002/aisy.202200104 |
| work_keys_str_mv | AT lingxiangjia explainablefragmentbasedmolecularpropertyattribution AT zunleifeng explainablefragmentbasedmolecularpropertyattribution AT haotianzhang explainablefragmentbasedmolecularpropertyattribution AT jiesong explainablefragmentbasedmolecularpropertyattribution AT zipengzhong explainablefragmentbasedmolecularpropertyattribution AT shaolunyao explainablefragmentbasedmolecularpropertyattribution AT minglisong explainablefragmentbasedmolecularpropertyattribution |