Share
Integrating AI-Driven Intelligent Agents and Multimodal Generative Models for Enhanced Cancer Diagnosis and Treatment
Doctoral study program
Biomedical Sciences, Molecular Medicine (Faculty of Medicine, Masaryk University)
Research area
Bioinformatics
Supervisor
Annotation
This Ph.D. project aims to revolutionize clinical oncology by integrating intelligent, AI-driven agents and multimodal generative artificial intelligence. The research will focus on developing and deploying advanced computational frameworks capable of processing, analyzing, and interpreting diverse biomedical datasets—including imaging (MRI, CT, PET), electronic health records, genetic, genomic, transcriptomic, proteomic, and epigenomic data.
The Ph.D. candidate will collaborate on the projects working towards the modular AI agents capable of securely operating within Trusted Research Environments (TREs). These agents will interface with multimodal foundational models, such as dnaBERT, epi-GPT, DeepSNP, and scGPT, deployed within these secure spaces. Agents will translate clinical queries into executable analytical workflows, orchestrate local data processing, and provide aggregated, interpretable outputs to researchers, ensuring compliance with ethical and privacy regulations.
The project will specifically explore the latent-space representations produced by foundational models to unify multimodal health data streams, thereby generating predictive and actionable insights to enhance patient stratification and clinical decision-making. Real-world use cases in hematology, cardiovascular diseases, triple-negative breast cancer, and prostate cancer will be leveraged for validation, particularly within projects like ACGT2, emphasizing long-read sequencing data integration.
Ultimately, this research will produce pioneering methodologies for multimodal generative AI-based predictive modeling, contributing to both clinical oncology and bioinformatics literature, and driving advancements in precision medicine.
Recommended literature
- Hao, M., Gong, J., Zeng, X., Liu, C., Guo, Y., Cheng, X., Wang, T., Ma, J., Song, L., & Zhang, X. (2023). “Large Scale Foundation Model on Single-cell Transcriptomics.” bioRxiv. https://doi.org/10.1101/2023.05.29.542705
- Wang, S., et al. (2023). “scGPT: leveraging GPT-like architecture for single-cell RNA-seq analysis.” Nature Methods.
- Wang, S., et al. (2020). “dnaBERT: Pre-trained Bidirectional Encoder Representations from Transformers model for DNA-language in genome research.” Nature Communications.
Funding
CEITEC Bioinformatics Core Facility budget, EOSC related projects, ELIXIR CZ projects
Requirements on candidates
Candidates should possess a solid background in bioinformatics, data science, machine learning, and computational biology, with demonstrated experience in multimodal health data integration and AI model development. Familiarity with transformer architectures (GPT models), TRE-based secure environments, and multiomics data analysis is highly desirable.
Keywords
Generative AI, Multimodal Data Integration, Intelligent Agents, Trusted Research Environments, Cancer Diagnosis, Bioinformatics, Foundational Models, Precision Medicine
Information on the supervisor
Number of successfully finished students: 1
Number of current students: 3
Number of current students over 4 years: 0