Magicmol: a light-weighted pipeline for drug-like molecule evolution and quick chemical space exploration
Abstract The flourishment of machine learning and deep learning methods has boosted the development of cheminformatics, especially regarding the application of drug discovery and new material exploration. Lower time and space expenses make it possible for scientists to search the enormous chemical s...
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
BMC
2023-04-01
|
| Series: | BMC Bioinformatics |
| Subjects: | |
| Online Access: | https://doi.org/10.1186/s12859-023-05286-0 |
| _version_ | 1797836252481847296 |
|---|---|
| author | Lin Chen Qing Shen Jungang Lou |
| author_facet | Lin Chen Qing Shen Jungang Lou |
| author_sort | Lin Chen |
| collection | DOAJ |
| description | Abstract The flourishment of machine learning and deep learning methods has boosted the development of cheminformatics, especially regarding the application of drug discovery and new material exploration. Lower time and space expenses make it possible for scientists to search the enormous chemical space. Recently, some work combined reinforcement learning strategies with recurrent neural network (RNN)-based models to optimize the property of generated small molecules, which notably improved a batch of critical factors for these candidates. However, a common problem among these RNN-based methods is that several generated molecules have difficulty in synthesizing despite owning higher desired properties such as binding affinity. However, RNN-based framework better reproduces the molecule distribution among the training set than other categories of models during molecule exploration tasks. Thus, to optimize the whole exploration process and make it contribute to the optimization of specified molecules, we devised a light-weighted pipeline called Magicmol; this pipeline has a re-mastered RNN network and utilize SELFIES presentation instead of SMILES. Our backbone model achieved extraordinary performance while reducing the training cost; moreover, we devised reward truncate strategies to eliminate the model collapse problem. Additionally, adopting SELFIES presentation made it possible to combine STONED-SELFIES as a post-processing procedure for specified molecule optimization and quick chemical space exploration. |
| first_indexed | 2024-04-09T15:06:55Z |
| format | Article |
| id | doaj.art-f4adafa4fb92440d9220ee33e090424a |
| institution | Directory Open Access Journal |
| issn | 1471-2105 |
| language | English |
| last_indexed | 2024-04-09T15:06:55Z |
| publishDate | 2023-04-01 |
| publisher | BMC |
| record_format | Article |
| series | BMC Bioinformatics |
| spelling | doaj.art-f4adafa4fb92440d9220ee33e090424a2023-04-30T11:30:11ZengBMCBMC Bioinformatics1471-21052023-04-0124111810.1186/s12859-023-05286-0Magicmol: a light-weighted pipeline for drug-like molecule evolution and quick chemical space explorationLin Chen0Qing Shen1Jungang Lou2Yangtze Delta Region (Huzhou) Institute of Intelligent Transportation, Huzhou UniversityYangtze Delta Region (Huzhou) Institute of Intelligent Transportation, Huzhou UniversityYangtze Delta Region (Huzhou) Institute of Intelligent Transportation, Huzhou UniversityAbstract The flourishment of machine learning and deep learning methods has boosted the development of cheminformatics, especially regarding the application of drug discovery and new material exploration. Lower time and space expenses make it possible for scientists to search the enormous chemical space. Recently, some work combined reinforcement learning strategies with recurrent neural network (RNN)-based models to optimize the property of generated small molecules, which notably improved a batch of critical factors for these candidates. However, a common problem among these RNN-based methods is that several generated molecules have difficulty in synthesizing despite owning higher desired properties such as binding affinity. However, RNN-based framework better reproduces the molecule distribution among the training set than other categories of models during molecule exploration tasks. Thus, to optimize the whole exploration process and make it contribute to the optimization of specified molecules, we devised a light-weighted pipeline called Magicmol; this pipeline has a re-mastered RNN network and utilize SELFIES presentation instead of SMILES. Our backbone model achieved extraordinary performance while reducing the training cost; moreover, we devised reward truncate strategies to eliminate the model collapse problem. Additionally, adopting SELFIES presentation made it possible to combine STONED-SELFIES as a post-processing procedure for specified molecule optimization and quick chemical space exploration.https://doi.org/10.1186/s12859-023-05286-0Generative modelsReinforcement learningDeep learningSynthetic accessibilityDe novo drug design |
| spellingShingle | Lin Chen Qing Shen Jungang Lou Magicmol: a light-weighted pipeline for drug-like molecule evolution and quick chemical space exploration BMC Bioinformatics Generative models Reinforcement learning Deep learning Synthetic accessibility De novo drug design |
| title | Magicmol: a light-weighted pipeline for drug-like molecule evolution and quick chemical space exploration |
| title_full | Magicmol: a light-weighted pipeline for drug-like molecule evolution and quick chemical space exploration |
| title_fullStr | Magicmol: a light-weighted pipeline for drug-like molecule evolution and quick chemical space exploration |
| title_full_unstemmed | Magicmol: a light-weighted pipeline for drug-like molecule evolution and quick chemical space exploration |
| title_short | Magicmol: a light-weighted pipeline for drug-like molecule evolution and quick chemical space exploration |
| title_sort | magicmol a light weighted pipeline for drug like molecule evolution and quick chemical space exploration |
| topic | Generative models Reinforcement learning Deep learning Synthetic accessibility De novo drug design |
| url | https://doi.org/10.1186/s12859-023-05286-0 |
| work_keys_str_mv | AT linchen magicmolalightweightedpipelinefordruglikemoleculeevolutionandquickchemicalspaceexploration AT qingshen magicmolalightweightedpipelinefordruglikemoleculeevolutionandquickchemicalspaceexploration AT junganglou magicmolalightweightedpipelinefordruglikemoleculeevolutionandquickchemicalspaceexploration |