Machine learning-based drug-drug interaction prediction: a critical review of models, limitations, and data challenges

Background/ObjectivesNew computational methods, based on statistical, machine learning, and deep learning techniques using drug-related entities (e.g., genes, protein bindings, etc.), help reduce the costs of in-vitro experiments through drug-drug interaction prediction (DDIp). This review examines...

Full description

Saved in:
Bibliographic Details
Main Authors: Flaviu-Ioan Gheorghita, Vlad-Ioan Bocanet, Laszlo Barna Iantovics
Format: Article
Language:English
Published: Frontiers Media S.A. 2025-07-01
Series:Frontiers in Pharmacology
Subjects:
Online Access:https://www.frontiersin.org/articles/10.3389/fphar.2025.1632775/full
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849411459474259968
author Flaviu-Ioan Gheorghita
Vlad-Ioan Bocanet
Laszlo Barna Iantovics
author_facet Flaviu-Ioan Gheorghita
Vlad-Ioan Bocanet
Laszlo Barna Iantovics
author_sort Flaviu-Ioan Gheorghita
collection DOAJ
description Background/ObjectivesNew computational methods, based on statistical, machine learning, and deep learning techniques using drug-related entities (e.g., genes, protein bindings, etc.), help reduce the costs of in-vitro experiments through drug-drug interaction prediction (DDIp). This review examines recent advances in DDIp. It presents an in-depth review of the state-of-the-art studies relating to semi-supervised, supervised, self-supervised learning, and other techniques such as graph-based learning and matrix factorization methods for predicting DDIs. All possible interactions between drugs are not known, and accurately predicting interactions is even more difficult due to the complex nature of drug-drug interactions (DDI).MethodsOf the 49 papers published in Web of Science in the last 6 years, 24 papers were considered relevant based on information presented in their titles and abstracts. The included articles focus specifically on predicting DDIs using a type of machine learning algorithm. Excluded articles focused on drug discovery, drug repurposing, molecular representation, or the extraction of biomedical interactions. The methodology, results limitations, and future research directions were studied for each paper. Common challenges, limitations, and future research directions were analyzed.Results and conclusionThe main limitations are class imbalance, poor performance on new drugs, limited explainability, and the need for additional data sources.
format Article
id doaj-art-fcfbf11dc1244e6796dc3efff597ddab
institution Kabale University
issn 1663-9812
language English
publishDate 2025-07-01
publisher Frontiers Media S.A.
record_format Article
series Frontiers in Pharmacology
spelling doaj-art-fcfbf11dc1244e6796dc3efff597ddab2025-08-20T03:34:45ZengFrontiers Media S.A.Frontiers in Pharmacology1663-98122025-07-011610.3389/fphar.2025.16327751632775Machine learning-based drug-drug interaction prediction: a critical review of models, limitations, and data challengesFlaviu-Ioan Gheorghita0Vlad-Ioan Bocanet1Laszlo Barna Iantovics2Doctoral School of Letters, Humanities and Applied Sciences, George Emil Palade University of Medicine, Pharmacy, Science, and Technology of Târgu Mureş, Târgu Mureş, RomaniaDepartment of Manufacturing Engineering, Technical University of Cluj-Napoca, Cluj-Napoca, RomaniaDepartment of Electrical Engineering and Information Technology, George Emil Palade University of Medicine, Pharmacy, Science, and Technology of Târgu Mureş, Târgu Mureş, RomaniaBackground/ObjectivesNew computational methods, based on statistical, machine learning, and deep learning techniques using drug-related entities (e.g., genes, protein bindings, etc.), help reduce the costs of in-vitro experiments through drug-drug interaction prediction (DDIp). This review examines recent advances in DDIp. It presents an in-depth review of the state-of-the-art studies relating to semi-supervised, supervised, self-supervised learning, and other techniques such as graph-based learning and matrix factorization methods for predicting DDIs. All possible interactions between drugs are not known, and accurately predicting interactions is even more difficult due to the complex nature of drug-drug interactions (DDI).MethodsOf the 49 papers published in Web of Science in the last 6 years, 24 papers were considered relevant based on information presented in their titles and abstracts. The included articles focus specifically on predicting DDIs using a type of machine learning algorithm. Excluded articles focused on drug discovery, drug repurposing, molecular representation, or the extraction of biomedical interactions. The methodology, results limitations, and future research directions were studied for each paper. Common challenges, limitations, and future research directions were analyzed.Results and conclusionThe main limitations are class imbalance, poor performance on new drugs, limited explainability, and the need for additional data sources.https://www.frontiersin.org/articles/10.3389/fphar.2025.1632775/fulldrug-drug interactionadverse drug reactionsmachine learning techniqueshealthcaresemi-supervised learningsupervised learning
spellingShingle Flaviu-Ioan Gheorghita
Vlad-Ioan Bocanet
Laszlo Barna Iantovics
Machine learning-based drug-drug interaction prediction: a critical review of models, limitations, and data challenges
Frontiers in Pharmacology
drug-drug interaction
adverse drug reactions
machine learning techniques
healthcare
semi-supervised learning
supervised learning
title Machine learning-based drug-drug interaction prediction: a critical review of models, limitations, and data challenges
title_full Machine learning-based drug-drug interaction prediction: a critical review of models, limitations, and data challenges
title_fullStr Machine learning-based drug-drug interaction prediction: a critical review of models, limitations, and data challenges
title_full_unstemmed Machine learning-based drug-drug interaction prediction: a critical review of models, limitations, and data challenges
title_short Machine learning-based drug-drug interaction prediction: a critical review of models, limitations, and data challenges
title_sort machine learning based drug drug interaction prediction a critical review of models limitations and data challenges
topic drug-drug interaction
adverse drug reactions
machine learning techniques
healthcare
semi-supervised learning
supervised learning
url https://www.frontiersin.org/articles/10.3389/fphar.2025.1632775/full
work_keys_str_mv AT flaviuioangheorghita machinelearningbaseddrugdruginteractionpredictionacriticalreviewofmodelslimitationsanddatachallenges
AT vladioanbocanet machinelearningbaseddrugdruginteractionpredictionacriticalreviewofmodelslimitationsanddatachallenges
AT laszlobarnaiantovics machinelearningbaseddrugdruginteractionpredictionacriticalreviewofmodelslimitationsanddatachallenges