Synergizing Multi-modal Agents and Large Foundation Models for Bioinformatics and Biomedicine

overview

The convergence of multi-modal agents and large foundation models is opening up transformative possibilities in bioinformatics and biomedicine. With the explosion of biological and biomedical data—from genomic sequences and proteomics to biomedical imaging and electronic health records—the need for intelligent systems that can effectively integrate and interpret this information has never been more critical. Multi-modal agents are designed to seamlessly process data across diverse modalities, while large foundation models, pre-trained on massive datasets, offer powerful generalization capabilities and adaptability across tasks.

This workshop aims to explore how the synergy between these two approaches can address long-standing challenges in the field, such as data heterogeneity, limited annotations, and model interpretability. By combining multi-modal agents with large foundation models, we can enable more robust disease gene prediction, biomarker discovery, drug response modeling, and other key tasks in computational biology and biomedicine.

The goal of this workshop is to bring together researchers and practitioners from machine learning, bioinformatics, and biomedical sciences to share insights, foster collaborations, and chart new directions. We are particularly interested in approaches that demonstrate:

Integrative Data Analysis: Using multi-modal strategies to connect and analyze omics data, imaging, and clinical records.

Knowledge Transfer: Leveraging foundation models to overcome data scarcity in biomedical domains.

Scalable and Trustworthy Systems:Building robust, interpretable tools for real-world biomedicine.

Novel Applications: Exploring innovative use cases in personalized medicine, drug discovery, and beyond.

Topics and Areas

This workshop invites contributions that explore the integration of multi-modal agents and large foundation models to advance research and applications in bioinformatics and biomedicine. Topics of interest include, but are not limited to:

Integration Techniques for Multi-modal Biological and Biomedical Data

Novel methods for fusing omics data, biomedical images, clinical records, and biosignals using large foundation models.
Techniques for aligning, synchronizing, and harmonizing heterogeneous biological data sources across modalities.

Adaptation of Foundation Models to Bioinformatics and Biomedical Domains

Transfer learning strategies for applying pre-trained models to tasks like genome analysis, proteomics, transcriptomics, and molecular phenotyping.
Domain-specific fine-tuning of foundation models for multi-modal biomedical applications.

Development of Intelligent Multi-modal Agents for Biomedicine

Design of AI agents that leverage foundation models to support decision-making in areas such as drug discovery, disease classification, and biomarker identification.
Applications of multi-modal agents in personalized medicine, precision health, and biomedical research.

Cross-modal and Multi-scale Representation Learning

Learning unified and interpretable representations across modalities (e.g., integrating genetic data with imaging or EHRs).
Techniques for linking molecular-level data with phenotype or population-level information through foundation models.

Biomedical Case Studies and Applications

Real-world case studies demonstrating the utility of combining multi-modal agents with large foundation models.
Evaluations of performance, interpretability, and impact in clinical and research settings.

Scalability, Efficiency, and Deployment

Computational strategies to scale foundation models for use in bioinformatics workflows and clinical pipelines.
Resource-efficient model deployment, including cloud-based and edge AI approaches for large-scale biomedical data.

Explainability and Trust in Biomedical AI

Frameworks to enhance transparency, reliability, and trustworthiness in multi-modal biomedical AI systems.
Visualization and interaction tools to support domain experts in interpreting model predictions.

Standardization, Benchmarks, and Evaluation Frameworks

Creation of benchmark datasets for evaluating multi-modal foundation model approaches in bioinformatics.
Comparative studies of model performance, robustness, and generalizability across tasks and data types.

Important Date:

Workshop paper submission deadline: 15 October 2025

Workshop paper notification: 10 November 2025

Workshop paper camera-ready: 23 November 2025

Workshop dates: 15-18 December 2025

Submission

Submit a New Paper

Papers should be formatted according to the IEEE Computer Society Proceedings Manuscript Formatting Guidelines. The template can be found via:

Important Notes:

Papers should be formatted to 8 pages IEEE Computer Society Proceedings Manuscript Formatting Guidelines.
You are strongly encouraged to print and double check your PDF file before its submission, especially if your paper contains Asian/European language symbols (such as Chinese/Korean characters or English letters with European fonts).

Transfer Policy:

If your paper was submitted but not accepted in the main conference, you cannot directly submit it to a workshop but it must be transferred into the workshop. You can change this option (workshop) at the conference paper submission site or (after the deadline) request the organizer of the workshop to which you want your paper to be transferred.

Submission Links:

Submit a new paper:

Keynote Speakers

To be Confirmed...

Workshop Organizers

Yanda Meng

University of Exeter, UK

Dr. Yanda Meng is an assistant professor (UK Lecturer) at the University of Exeter, Computer Science Department. His research is mainly at the interface of artificial intelligence and healthcare, specialised in the research and development of artificial intelligence, biomedical image processing and analysis techniques. He has published more than 40 papers in peer-reviewed journals and conferences such as IEEE Transactions on Medical Imaging, Medical Image Analysis, Diabetologia, Cardiovascular Diabetology, CVPR, ICCV, ECCV, AAAI, and MICCAI, etc. He has leadership roles as project PI delivering successful projects working effectively with different stakeholders (academics, clinicians, public, and industry). His collaboration with clinicians has made significant impacts on patient care. His research has been or is being supported by UKRI, UK Royal Society, Exeter Innovation, NHS Foundation Trust, and Welcome Trust. Yanda has also served as the Topic Editor and Guest Editor for many journals, such as IEEE JBHI, Frontiers in Medicine, Mathematics; the Area Chair for MICCAI 2025 and IEEE ICARM 2025; the Program Committee Board member for ISBI 2026.

Wei Zhou

Cardiff University, UK

Dr. Wei Zhou is an Assistant Professor (UK Lecturer) at Cardiff University. Wei’s research interests mainly focus on perceptual image processing, multimodality, and visual computing for healthcare. Dr Zhou has published over 70 papers in recent years, including publications in top-tier venues, e.g., IEEE TIP, IEEE TMM, IEEE TCSVT, IEEE TMI, CVPR, ACM MM, MICCAI, etc. Wei serves as General Chair for the 1st Cardiff Image & Vision Computing Workshop and Chair for the Elections Committee of IEEE UK & Ireland SPS Chapter. Wei is now an Associate Editor of IEEE Transactions on Neural Networks and Learning Systems (TNNLS), Pattern Recognition, Neurocomputing, Springer Signal, Image and Video Processing, and Human-centric Computing and Information Sciences. Wei has also served as the Topic Editor and Guest Editor for many journals, such as Elsevier Displays; the Area Chair for ACM MM 2024, ICME 2025, and IJCNN 2025; the Lead Special Session Chair for IEEE ICME 2025, IEEE QoMEX 2025, IEEE MMSP 2023, and the Special Session Co-Chair for IEEE ICIP 2025 and 2024.

Meng Wang

National University of Singapore, Singapore

Dr Meng Wang is a Senior Research Fellow at the Centre for Innovation and Precision Eye Health, National University of Singapore. His research focuses on medical artificial intelligence and multimodal medical image analysis, specifically emphasizing medical image analysis, large foundation models for medical imaging, and trustworthy AI. Dr. Meng Wang has published over 50 papers, including in top international journals like Nature Communications, npj Digital Medicine, Cell Reports Medicine, IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI), and IEEE Transactions on Medical Imaging (TMI), as well as at leading international conferences such as CVPR and MICCAI. He has also contributed to the authorship of the specialized book "Federated Learning for Medical Imaging.

Qinkai Yu

University of Exeter, UK

Mr Qinkai Yu is a PhD student in the Department of Computer Science at the University of Exeter, under the supervision of Dr. Yanda Meng. His research lies at the intersection of medical image analysis and the application of large language models and multimodal foundation models in healthcare. He has published multiple papers as the first author or co-first author in top-tier peer-reviewed conferences such as ACL, COLING, and MICCAI. In 2025, two of his first-author papers have been accepted early by MICCAI.