Molecular image-based convolutional neural network for the prediction of ADMET properties

Shi, Tingting; Yang, Yingwu; Huang, Shuheng; Chen, Linxin; Kuang, Zuyin; Heng, Yu; Mei, Hu

doi:10.1016/j.chemolab.2019.103853

Cited by 46 publications

(27 citation statements)

References 26 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Notably, Kumar et al [135] used 3D-RISM-KH based solvation free energy descriptors to increase performance. Meanwhile, Shi et al [136] used CNN to extract a task-specific feature of the molecule and predicted its four ADMET properties from the 2D structure image. The dataset contains CYP1A2 inhibitors, P-gp inhibitors, BBB penetrating agents, and Ames mutagens.…”

Section: Distributionmentioning

confidence: 99%

“…At present, many researchers are attempting to make accurate prediction models and find structural patterns. Both machine learningbased [137][138][139] and deep learning-based models [136,140] for BBB prediction appear currently. While Toropov et al…”

Section: Distributionmentioning

confidence: 99%

“…Thus, it partitions label space and trains the base classifier to classify each subspace separately. Besides the aforementioned substrate prediction studies, other remarkable CYP450 inhibitor prediction studies have been published [136,[144][145][146]. Pang et al [145] collected data from BindingDB and ChEMBL and constructed a CYP450 3A4 isoform inhibitor prediction model.…”

Section: Metabolism and Excretionmentioning

confidence: 99%

“…Specifically, most predictive models comprise hundreds to thousands of small chemistry datasets that cannot cover enough chemical space [76,118,188]. Moreover, the data is usually dispersed to many literatures [117,118,122,124,125,[128][129][130][134][135][136][137][138][139][140][141][142][143][144]150,155,[159][160][161][162][163][164][165][166][167][168][169]173,176,179,183], is unbalanced, and has cutoff ambiguity challenges [118,167]. Furthermore, the bioactivity assay data is strongly biased to its platform, and has an intrinsic experimental error which disrupts accurate prediction [189].…”

Section: Limitations and Future Directionmentioning

confidence: 99%

“…Second, most of the studies used engineered molecular descriptors to train the conventional machine learning models, but there are feature-intrinsic biases and a modelinherent low interpretability [72]. To address these problems, a novel data-driven feature generation [71,73] or targetspecific feature learning from raw data offer good solutions [76,136,154,179,184,185]. However, having interpretability in the conventional machine learning and deep learning model is still challenging.…”

Section: Limitations and Future Directionmentioning

confidence: 99%

See 4 more Smart Citations

Artificial Intelligence in Drug Discovery: A Comprehensive Review of Data-driven and Machine Learning Approaches

Kim

Lee

et al. 2020

Biotechnol Bioproc E

View full text Add to dashboard Cite

Section: Distributionmentioning

confidence: 99%

Section: Distributionmentioning

confidence: 99%

Section: Metabolism and Excretionmentioning

confidence: 99%

Section: Limitations and Future Directionmentioning

confidence: 99%

Section: Limitations and Future Directionmentioning

confidence: 99%

See 3 more Smart Citations

Artificial Intelligence in Drug Discovery: A Comprehensive Review of Data-driven and Machine Learning Approaches

Kim

Lee

et al. 2020

Biotechnol Bioproc E

View full text Add to dashboard Cite

Explainable Fragment‐Based Molecular Property Attribution

Jia

Feng

Zhang

et al. 2022

Advanced Intelligent Systems

View full text Add to dashboard Cite

“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.

show abstract

Deep Neural Networks for QSAR

2021

Artificial Intelligence in Drug Design

View full text Add to dashboard Cite

Molecular image-based convolutional neural network for the prediction of ADMET properties

Cited by 46 publications

References 26 publications

Artificial Intelligence in Drug Discovery: A Comprehensive Review of Data-driven and Machine Learning Approaches

Artificial Intelligence in Drug Discovery: A Comprehensive Review of Data-driven and Machine Learning Approaches

Explainable Fragment‐Based Molecular Property Attribution

Deep Neural Networks for QSAR

Contact Info

Product

Resources

About