Apr 2026· Journal of Chemical Information and Modeling· Vol 66 9, pp.
5008-5017
· 10 citations· ⚡ 1 influential· 56 references
Computer ScienceMedicine
TL;DR
The DRHIN platform provides a code-free portal supporting three key predictive tasks: discovering drug-disease associations, repurposing existing drugs for new indications, and identifying potential therapies for specific diseases, making analyses accessible and reproducible.
Abstract
Drug repositioning (DR) identifies new therapeutic uses for approved drugs, reducing development burdens and offering safer treatment options for patients. While high-throughput technologies generate complex, large-scale multiomics data, existing DR tools struggle to comprehensively analyze the resulting biological networks. To address this challenge, we present DRHIN, an integrated, interactive web server for DR over heterogeneous information networks (HINs) using advanced deep learning techniques. DRHIN integrates transcriptomics, proteomics, and microbiome data, incorporating eight biological entities and 19 association types to build diverse HINs and elucidate the underlying molecular mechanisms. It includes 19 state-of-the-art graph representation algorithms, enabling flexible training, comparison, and evaluation of heterogeneous network data. The platform provides a code-free portal supporting three key predictive tasks: discovering drug-disease associations, repurposing existing drugs for new indications, and identifying potential therapies for specific diseases, making analyses accessible and reproducible. Leveraging high-performance computing, DRHIN efficiently processes million-scale networks, ensuring practical applicability in real-world scenarios. The web server is freely accessible at http://drhin.tianshanzw.cn.
MKASynergy, an adaptive drug synergy prediction method based on a mixture-of-experts kernel mechanism, achieves competitive predictive performance and visualization analysis confirms the model’s effectiveness in feature decoupling and helps interpret latent drug synergistic mechanisms.
Cundong Lin, Jiancheng Ni, Ying Yang et al.· Network Modeling Analysis in...· 0 citations
Drug-drug interaction (DDI) event prediction is critical for ensuring patient safety and optimizing therapeutic outcomes. Existing computational approaches are limited by their inability to jointly model the heterogeneous mechanisms underlying DDIs, which span molecular structure, pharmacodynamic function, and network-mediated relations. To address this limitation, we introduce M2DDI, a unified framework for dynamic multimodal fusion in DDI prediction. M2DDI utilizes a Mixture-of-Experts architecture, with each expert dedicated to a distinct pharmacological modality. A novel prior-enhanced dual-path gating strategy adaptively selects relevant experts for each drug pair by integrating mechanism-matched feature queries and ATC-based biomedical priors, thereby aligning expert selection with underlying pharmacological mechanisms and addressing the challenge of data incompleteness. Empirical evaluation on benchmark datasets demonstrates that M2DDI achieves state-of-the-art performance, particularly in new drug scenarios. Additional robustness experiments show that M2DDI maintains high predictive accuracy even when modality-specific information is partially missing, outperforming existing methods under similar conditions. Analysis of expert selection patterns further confirms alignment with established pharmacological mechanisms. These results establish M2DDI as an effective and mechanism-aware solution for comprehensive DDI prediction. The code is available at: https://github.com/RunqingXuCn/M2DDI.
Runqing Xu, Siyi Liu, Haoyang Li et al.· Proceedings of the 32nd ACM...· 0 citations
IHLO-DTI, a novel prediction model based on an improved hypergraph neural network and Laplacian matrix optimization, can effectively capture high-order many-to-many interactions between drugs and targets, improving prediction accuracy and robustness.
Guolongwei Dai, Tao Luo, Dandan Li et al.· ACS Synthetic Biology· 0 citations
Drug-drug interactions (DDIs) represent a substantial challenge in contemporary
pharmacotherapy, especially given polypharmacy, the effects of foods, and the modification
of host-microbiota systems on drugs. Although useful, existing DDI identification techniques have
many constraints related to cost, time, and scalability.
A literature review was conducted using PubMed, Scopus, Web of Science, and IEEE
Xplore, focusing on machine learning techniques, deep neural architectures, and network-based
models that integrate multi-omic, pharmacological, and clinical data.
By combining chemical, biological, and clinical data into scalable computer platforms,
demonstrated that artificial intelligence techniques, such as machine learning (ML) and deep learning
(DL), are changing the prediction of DDI. Some notable studies, such as DeepDDI, TP-DDI, and
Decagon, use approaches that successfully capture the intricate PK-PD interactions of pharmaceuticals.
On the other hand, food-drug interactions and microbiome-mediated drug interactions were also
successfully predicted using multimodal and graph-based models, respectively.
Critical issues, such as insufficient data, class imbalance, and model interpretability,
must be addressed through explainable AI and multimodal fusion techniques.
The purpose of this article is to present an overview of how artificial intelligence might
serve not only as a tool but also as a strategic solution for safe prescribing and tailored pharmacotherapy,
hence opening up new avenues for the field of drug safety science.
D. Tripathi, Ankita Wal, Vivek Kumar Gupta et al.· Current Bioinformatics· 0 citations
The key innovation of LOGIC lies in constructing a dictionary of functional groups and symptoms, and performing a simple and intuitive multi-hot encoding of drugs and diseases at the micro-scale, and in employing large language models (LLMs) to derive the meso-scale features of diseases without requiring additional domain knowledge.
Yunfei He, Shikai Chen, Yuchen Zhao et al.· IEEE transactions on computa...· 0 citations
Druggable proteins are proteins that can be specifically bound and modulated by drug molecules, with such modulation expected to produce therapeutic effects. The identification and validation of druggable proteins are central steps in modern drug discovery. With the rapid advancements in biological big data and artificial intelligence, computational methods based on bioinformatics have become essential for large-scale screening of druggable proteins. This review systematically summarizes the latest research progress in this field, providing a comprehensive analysis across three dimensions: data resources, feature engineering, and predictive models. The main contributions are 4-fold: First, we systematically describe three major types of databases: pharmacological annotation databases, quantitative affinity databases, and multi-source integrated databases, and dataset construction strategies as well as strategies for handling imbalanced samples are explored. Second, multi-level feature representations across four dimensions are categorized into sequencederived features, physicochemical properties, evolutionary information, and structural information, and the corresponding methods and tools are summarized. Third, we provide a detailed review of the evolution and application paradigms of predictive models, from classical machine learning and ensemble learning to deep learning architectures and protein large language models. Fourth, based on a systematic evaluation of representative studies, current limitations in areas such as few-shot learning and model interpretability are highlighted, and offer a forward-looking perspective on future research directions. This review aims to serve as a comprehensive reference for druggable protein prediction algorithms and to provide a solid theoretical and technical roadmap for computationally driven discovery of novel drug targets.
Hong-Qi Zhang, Hong-Ling Wang, Shang-Hua Liu et al.· Current Drug Targets· 0 citations
Related blog posts
MIT News · Artificial Intelligence· news.mit.eduAug 3, 2026
A new machine-learning framework aims to improve the success rate of computational protein design while moving away from results that reproduce sequences found in nature.
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.
