Cold-start problems in data-driven prediction of drug-drug interaction effects
- Author
- Pieter Dewulf (UGent) , Michiel Stock (UGent) and Bernard De Baets (UGent)
- Organization
- Project
- Abstract
- Combining drugs, a phenomenon often referred to as polypharmacy, can induce additional adverse effects. The identification of adverse combinations is a key task in pharmacovigilance. In this context, in silico approaches based on machine learning are promising as they can learn from a limited number of combinations to predict for all. In this work, we identify various subtasks in predicting effects caused by drug–drug interaction. Predicting drug–drug interaction effects for drugs that already exist is very different from predicting outcomes for newly developed drugs, commonly called a cold-start problem. We propose suitable validation schemes for the different subtasks that emerge. These validation schemes are critical to correctly assess the performance. We develop a new model that obtains AUC-ROC =0.843 for the hardest cold-start task up to AUC-ROC =0.957 for the easiest one on the benchmark dataset of Zitnik et al. Finally, we illustrate how our predictions can be used to improve post-market surveillance systems or detect drug–drug interaction effects earlier during drug development.
- Keywords
- Molecular Medicine, Pharmaceutical Science, polypharmacy, drug-drug interaction, prediction, cross-validation, machine learning, cold-start problems, UNITED-STATES, MECHANISMS, SAFETY, ADULTS, TRENDS
Downloads
-
KERMIT-A1-623.pdf
- full text (Published version)
- |
- open access
- |
- |
- 1.33 MB
Citation
Please use this url to cite or link to this publication: http://hdl.handle.net/1854/LU-8707774
- MLA
- Dewulf, Pieter, et al. “Cold-Start Problems in Data-Driven Prediction of Drug-Drug Interaction Effects.” PHARMACEUTICALS, vol. 14, no. 5, 2021, doi:10.3390/ph14050429.
- APA
- Dewulf, P., Stock, M., & De Baets, B. (2021). Cold-start problems in data-driven prediction of drug-drug interaction effects. PHARMACEUTICALS, 14(5). https://doi.org/10.3390/ph14050429
- Chicago author-date
- Dewulf, Pieter, Michiel Stock, and Bernard De Baets. 2021. “Cold-Start Problems in Data-Driven Prediction of Drug-Drug Interaction Effects.” PHARMACEUTICALS 14 (5). https://doi.org/10.3390/ph14050429.
- Chicago author-date (all authors)
- Dewulf, Pieter, Michiel Stock, and Bernard De Baets. 2021. “Cold-Start Problems in Data-Driven Prediction of Drug-Drug Interaction Effects.” PHARMACEUTICALS 14 (5). doi:10.3390/ph14050429.
- Vancouver
- 1.Dewulf P, Stock M, De Baets B. Cold-start problems in data-driven prediction of drug-drug interaction effects. PHARMACEUTICALS. 2021;14(5).
- IEEE
- [1]P. Dewulf, M. Stock, and B. De Baets, “Cold-start problems in data-driven prediction of drug-drug interaction effects,” PHARMACEUTICALS, vol. 14, no. 5, 2021.
@article{8707774,
abstract = {{Combining drugs, a phenomenon often referred to as polypharmacy, can induce additional adverse effects. The identification of adverse combinations is a key task in pharmacovigilance. In this context, in silico approaches based on machine learning are promising as they can learn from a limited number of combinations to predict for all. In this work, we identify various subtasks in predicting effects caused by drug–drug interaction. Predicting drug–drug interaction effects for drugs that already exist is very different from predicting outcomes for newly developed drugs, commonly called a cold-start problem. We propose suitable validation schemes for the different subtasks that emerge. These validation schemes are critical to correctly assess the performance. We develop a new model that obtains AUC-ROC =0.843 for the hardest cold-start task up to AUC-ROC =0.957 for the easiest one on the benchmark dataset of Zitnik et al. Finally, we illustrate how our predictions can be used to improve post-market surveillance systems or detect drug–drug interaction effects earlier during drug development.}},
articleno = {{429}},
author = {{Dewulf, Pieter and Stock, Michiel and De Baets, Bernard}},
issn = {{1424-8247}},
journal = {{PHARMACEUTICALS}},
keywords = {{Molecular Medicine,Pharmaceutical Science,polypharmacy,drug-drug interaction,prediction,cross-validation,machine learning,cold-start problems,UNITED-STATES,MECHANISMS,SAFETY,ADULTS,TRENDS}},
language = {{eng}},
number = {{5}},
pages = {{23}},
title = {{Cold-start problems in data-driven prediction of drug-drug interaction effects}},
url = {{http://doi.org/10.3390/ph14050429}},
volume = {{14}},
year = {{2021}},
}
- Altmetric
- View in Altmetric
- Web of Science
- Times cited: