Large-scale comparison of machine learning methods for profiling prediction of kinase inhibitors
Abstract Conventional machine learning (ML) and deep learning (DL) play a key role in the selectivity prediction of kinase inhibitors. A number of models based on available datasets can be used to predict the kinase profile of compounds, but there is still controversy about the advantages and disadv...
Main Authors: | , , , , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
BMC
2024-01-01
|
Series: | Journal of Cheminformatics |
Subjects: | |
Online Access: | https://doi.org/10.1186/s13321-023-00799-5 |
_version_ | 1797273445736644608 |
---|---|
author | Jiangxia Wu Yihao Chen Jingxing Wu Duancheng Zhao Jindi Huang MuJie Lin Ling Wang |
author_facet | Jiangxia Wu Yihao Chen Jingxing Wu Duancheng Zhao Jindi Huang MuJie Lin Ling Wang |
author_sort | Jiangxia Wu |
collection | DOAJ |
description | Abstract Conventional machine learning (ML) and deep learning (DL) play a key role in the selectivity prediction of kinase inhibitors. A number of models based on available datasets can be used to predict the kinase profile of compounds, but there is still controversy about the advantages and disadvantages of ML and DL for such tasks. In this study, we constructed a comprehensive benchmark dataset of kinase inhibitors, involving in 141,086 unique compounds and 216,823 well-defined bioassay data points for 354 kinases. We then systematically compared the performance of 12 ML and DL methods on the kinase profiling prediction task. Extensive experimental results reveal that (1) Descriptor-based ML models generally slightly outperform fingerprint-based ML models in terms of predictive performance. RF as an ensemble learning approach displays the overall best predictive performance. (2) Single-task graph-based DL models are generally inferior to conventional descriptor- and fingerprint-based ML models, however, the corresponding multi-task models generally improves the average accuracy of kinase profile prediction. For example, the multi-task FP-GNN model outperforms the conventional descriptor- and fingerprint-based ML models with an average AUC of 0.807. (3) Fusion models based on voting and stacking methods can further improve the performance of the kinase profiling prediction task, specifically, RF::AtomPairs + FP2 + RDKitDes fusion model performs best with the highest average AUC value of 0.825 on the test sets. These findings provide useful information for guiding choices of the ML and DL methods for the kinase profiling prediction tasks. Finally, an online platform called KIPP ( https://kipp.idruglab.cn ) and python software are developed based on the best models to support the kinase profiling prediction, as well as various kinase inhibitor identification tasks including virtual screening, compound repositioning and target fishing. |
first_indexed | 2024-03-07T14:43:23Z |
format | Article |
id | doaj.art-20548c7c94374ecb8f669bdba07aed21 |
institution | Directory Open Access Journal |
issn | 1758-2946 |
language | English |
last_indexed | 2024-03-07T14:43:23Z |
publishDate | 2024-01-01 |
publisher | BMC |
record_format | Article |
series | Journal of Cheminformatics |
spelling | doaj.art-20548c7c94374ecb8f669bdba07aed212024-03-05T20:06:10ZengBMCJournal of Cheminformatics1758-29462024-01-0116112210.1186/s13321-023-00799-5Large-scale comparison of machine learning methods for profiling prediction of kinase inhibitorsJiangxia Wu0Yihao Chen1Jingxing Wu2Duancheng Zhao3Jindi Huang4MuJie Lin5Ling Wang6Guangdong Provincial Key Laboratory of Fermentation and Enzyme Engineering, Joint International Research Laboratory of Synthetic Biology and Medicine, Guangdong Provincial Engineering and Technology Research Center of Biopharmaceuticals, School of Biology and Biological Engineering, South China University of TechnologyGuangdong Provincial Key Laboratory of Fermentation and Enzyme Engineering, Joint International Research Laboratory of Synthetic Biology and Medicine, Guangdong Provincial Engineering and Technology Research Center of Biopharmaceuticals, School of Biology and Biological Engineering, South China University of TechnologyGuangdong Provincial Key Laboratory of Fermentation and Enzyme Engineering, Joint International Research Laboratory of Synthetic Biology and Medicine, Guangdong Provincial Engineering and Technology Research Center of Biopharmaceuticals, School of Biology and Biological Engineering, South China University of TechnologyGuangdong Provincial Key Laboratory of Fermentation and Enzyme Engineering, Joint International Research Laboratory of Synthetic Biology and Medicine, Guangdong Provincial Engineering and Technology Research Center of Biopharmaceuticals, School of Biology and Biological Engineering, South China University of TechnologyGuangdong Provincial Key Laboratory of Fermentation and Enzyme Engineering, Joint International Research Laboratory of Synthetic Biology and Medicine, Guangdong Provincial Engineering and Technology Research Center of Biopharmaceuticals, School of Biology and Biological Engineering, South China University of TechnologyGuangdong Provincial Key Laboratory of Fermentation and Enzyme Engineering, Joint International Research Laboratory of Synthetic Biology and Medicine, Guangdong Provincial Engineering and Technology Research Center of Biopharmaceuticals, School of Biology and Biological Engineering, South China University of TechnologyGuangdong Provincial Key Laboratory of Fermentation and Enzyme Engineering, Joint International Research Laboratory of Synthetic Biology and Medicine, Guangdong Provincial Engineering and Technology Research Center of Biopharmaceuticals, School of Biology and Biological Engineering, South China University of TechnologyAbstract Conventional machine learning (ML) and deep learning (DL) play a key role in the selectivity prediction of kinase inhibitors. A number of models based on available datasets can be used to predict the kinase profile of compounds, but there is still controversy about the advantages and disadvantages of ML and DL for such tasks. In this study, we constructed a comprehensive benchmark dataset of kinase inhibitors, involving in 141,086 unique compounds and 216,823 well-defined bioassay data points for 354 kinases. We then systematically compared the performance of 12 ML and DL methods on the kinase profiling prediction task. Extensive experimental results reveal that (1) Descriptor-based ML models generally slightly outperform fingerprint-based ML models in terms of predictive performance. RF as an ensemble learning approach displays the overall best predictive performance. (2) Single-task graph-based DL models are generally inferior to conventional descriptor- and fingerprint-based ML models, however, the corresponding multi-task models generally improves the average accuracy of kinase profile prediction. For example, the multi-task FP-GNN model outperforms the conventional descriptor- and fingerprint-based ML models with an average AUC of 0.807. (3) Fusion models based on voting and stacking methods can further improve the performance of the kinase profiling prediction task, specifically, RF::AtomPairs + FP2 + RDKitDes fusion model performs best with the highest average AUC value of 0.825 on the test sets. These findings provide useful information for guiding choices of the ML and DL methods for the kinase profiling prediction tasks. Finally, an online platform called KIPP ( https://kipp.idruglab.cn ) and python software are developed based on the best models to support the kinase profiling prediction, as well as various kinase inhibitor identification tasks including virtual screening, compound repositioning and target fishing.https://doi.org/10.1186/s13321-023-00799-5Kinase profilingMachine learningDeep learningMolecular fingerprintsMolecular graphs |
spellingShingle | Jiangxia Wu Yihao Chen Jingxing Wu Duancheng Zhao Jindi Huang MuJie Lin Ling Wang Large-scale comparison of machine learning methods for profiling prediction of kinase inhibitors Journal of Cheminformatics Kinase profiling Machine learning Deep learning Molecular fingerprints Molecular graphs |
title | Large-scale comparison of machine learning methods for profiling prediction of kinase inhibitors |
title_full | Large-scale comparison of machine learning methods for profiling prediction of kinase inhibitors |
title_fullStr | Large-scale comparison of machine learning methods for profiling prediction of kinase inhibitors |
title_full_unstemmed | Large-scale comparison of machine learning methods for profiling prediction of kinase inhibitors |
title_short | Large-scale comparison of machine learning methods for profiling prediction of kinase inhibitors |
title_sort | large scale comparison of machine learning methods for profiling prediction of kinase inhibitors |
topic | Kinase profiling Machine learning Deep learning Molecular fingerprints Molecular graphs |
url | https://doi.org/10.1186/s13321-023-00799-5 |
work_keys_str_mv | AT jiangxiawu largescalecomparisonofmachinelearningmethodsforprofilingpredictionofkinaseinhibitors AT yihaochen largescalecomparisonofmachinelearningmethodsforprofilingpredictionofkinaseinhibitors AT jingxingwu largescalecomparisonofmachinelearningmethodsforprofilingpredictionofkinaseinhibitors AT duanchengzhao largescalecomparisonofmachinelearningmethodsforprofilingpredictionofkinaseinhibitors AT jindihuang largescalecomparisonofmachinelearningmethodsforprofilingpredictionofkinaseinhibitors AT mujielin largescalecomparisonofmachinelearningmethodsforprofilingpredictionofkinaseinhibitors AT lingwang largescalecomparisonofmachinelearningmethodsforprofilingpredictionofkinaseinhibitors |