A comprehensive landscape of AI applications in broad-spectrum drug interaction prediction: a systematic review

Abstract

In drug development, managing interactions such as drug-drug, drug-disease, and drug-nutrient is critical for ensuring the safety and efficacy of pharmacological treatments. These interactions often overlap, forming a complex, interconnected landscape that necessitates accurate prediction to improve patient outcomes and support evidence-based care. Recent advances in artificial intelligence (AI), powered by large-scale datasets (e.g., DrugBank, TWOSIDES, SIDER), have significantly enhanced interaction prediction. Machine learning, deep learning, and graph-based models show great promise, but challenges persist, including data imbalance, noisy sources, Limited explainability, and underrepresentation of certain types of interactions. This systematic review of 147 studies (2018-2024) is the first to comprehensively map AI applications across major interaction types. We present a detailed taxonomy of models and datasets, emphasizing the growing roles of large language models and knowledge graphs in overcoming key limitations. Their integration-alongside explainable AI tools-enhances transparency, paving the way for AI-driven systems that proactively mitigate adverse interactions. By identifying the most promising approaches and critical research gaps, this review lays the groundwork for advancing more robust, interpretable, and personalized models for drug interaction prediction.

Keywords: Artificial intelligence; Deep learning; Drug–disease interactions; Drug–drug interactions; Drug–nutrient interactions; Graph-based method; Machine learning; Multiple drug interactions.

Fig. 1

a Schematic representation of the six major categories of drug interactions and their interconnected relationships. b Venn diagram showing the distribution of dataset availability (left) and artificial intelligence model applications (right) across the three primary interaction types. c A flowchart of drug interaction prediction (drug–drug, drug–-disease, and drug–nutrient) utilizing various data sources, machine learning, deep learning, and graph-based methods for safety evaluation. Created with BioRender.com

Fig. 2

PRISMA chart of search results for every category of drug interaction.

Fig. 3

Representative case studies illustrating different types of drug interactions, including drug–drug, drug–disease, drug-allergy, drug–food, drug–supplement, and drug–microbiome interactions, each highlighting distinct mechanisms and clinical consequences.

Fig. 4

Chronological visualization (2018–2024) of AI approaches in DDIs research, showing the evolution of machine learning (top), deep learning (right), and graph-based method (bottom). Bubble size reflects dataset size; colors represent addressed challenges. Years progress from bottom to top within each method.

Fig. 5

Sunburst plot of common datasets and AI models used across different drug interaction types, with sizes proportional to their frequency of occurrence in the literature.

Fig. 6

a Heatmap showing the number of studies investigating each drug interaction type using different AI model categories. b Heatmap displaying the average values of key performance metrics reported for each drug interaction type and AI model category.

Fig. 7

Alluvial plot illustrating the relationships between drug interaction types, AI model types, and utilized feature categories in predictive modeling

Fig. 8

Future directions in AI-driven drug interaction prediction. Created with BioRender.com