The project will explore the potentially transformative value of Foundation Models (FM), generative AI (GenAI) and Agentic Workflows for accelerating scientific discovery in biological sciences, particularly in the field of glycomics. These emerging in-silico technologies have already given rise to models, such as AlphaFold, which aim to accelerate research in structural biology, drug discovery, and disease understanding. The complexity of glycomics arises from the structural diversity, branching patterns, dynamic modifications, and functional versatility of glycans. These glycans affect cell-cell communication, immune system regulation, protein folding, pathogen recognition, signal recognition and have important implications across biological systems.
Current experimental and computational methods continue to generate vast amounts of biological and glycan-related data, yet their integration and analysis remain a bottleneck in translating this information into actionable biological insights and new hypotheses. When combined, these technologies promise to handle large-scale data from multiple sources (e.g., knowledge bases, experiment databases) to learn complex patterns and carry out inferences. Additionally, the development of novel agentic tools and workflows can further bring these technologies together and unlock new research avenues. By iteratively probing multiple models, knowledge, and tools, agents may be used to design and test new hypotheses, perform and validate in-silico experiments, or suggest the most promising experiments to be designed and undertaken. These activities will generate significant novel insight into the fields of Immunology and Neuroscience, progress novel target identification and (bio)therapeutic design as well as advance computational methods and GenAI.
The student will focus on building bridges between these technologies and domains to advance the fields of glycobiology and glycoinformatics, bioinformatics and computer science.
The student will have access to rich datasets from cutting-edge facilities as well as IBM's compute infrastructure.
We offer a fully funded 4-year PhD position at the intersection between computer systems design and AI as part of a new initiative between IBM Research Europe and Trinity College Dublin (TCD).
The PhD project will be jointly supervised by Professor Gavin Davey, School of Biochemistry and Immunology, Trinity College Dublin and Dr Joao Bettencourt-Silva (IBM Research Europe).
The selected student will be employed by IBM for the duration of the PhD and be a registered student at TCD with the following benefits:
- Access to resources and expertise both at IBM Research and TCD
- Research experience in both private and public sectors
- A substantial PhD Salary (>40,000 euro per annum)
- Full TCD PhD program fees (EU or non-EU level)
The following research questions will be tackled by the student:
- How can emerging multimodal foundation models (e.g., for small molecules, proteins, genes), LLMs (e.g., Granite) and other domain-specific models (e.g., AlphaFold) be used together with existing tools and knowledge bases (e.g., Glycologue, UniCarb-DB) to create new, more detailed models, of glycoproteins and advance glycoprotein research?
- How may agentic workflows be developed to integrate heterogeneous experimental datasets, perform iterative hypothesis generation, and guide experimental designs?
- Can models be trained to identify novel glycan targets in human diseases like cancer and optimise antibody design and synthesis?
- How do we use agentic workflows to model glycosyltransferase and glycosidase networks and train them to generate novel glycan structures which have nutritional and therapeutic benefits?
- Investigate how agentic workflows, computational tools and models may be applied and scaled across domains within scientific discovery workflows.
Examples of models, tools and databases and references:
- Glycologue https://glycologue.org
- UniCarb-DB https://unicarb-db.expasy.org
- Alphafold https://www.nature.com/articles/s41586-024-07487-w
- Multi-view Molecular Embedding with Late Fusion (MMELON) https://arxiv.org/abs/2410.19704
- Molecular Aligned Multi-Modal Architecture and Language (MAMMAL) https://arxiv.org/abs/2410.22367
- Otter-Knowledge: benchmarks of multimodal knowledge graph representation learning from different sources for drug discovery https://arxiv.org/abs/2306.12802
- The Virtual Lab: AI Agents Design New SARS-CoV-2 Nanobodies with Experimental Validation https://www.biorxiv.org/content/10.1101/2024.11.11.623004v1
- Enhancing foundation models for scientific discovery via multimodal knowledge graph representations https://www.sciencedirect.com/science/article/pii/S1570826824000313
- An honours degree in a relevant domain.
- Master’s Degree in one of the relevant fields such as Bioinformatics, Computational Biology, Data Analytics, Computer Science.
- Basic knowledge of Machine Learning and Artificial Intelligence: ML fundamentals and common Deep Learning architectures; experience with AI/ML libraries such as Pytorch.
- Good programming skills: Python, version control, ability to write complex code from scratch and extend existing codes.
- Strong interest in performing basic research work, including: evaluation of existing literature, ideation of possible/alternative solutions, writing research papers.
- Ability to move quickly from idea to software prototype for evaluation and demonstration.
- Excellent spoken and written English.
- An honours degree in biological sciences, biochemistry, bioinformatics or similar.
- A background in computational and/or data-driven methods.
- Experience applying computational methods to scientific or biological data.
- Knowledge of foundation model architectures with a specific focus on models using biological data and multi-modality.
- Knowledge of Agentic Workflows, common architectures and frameworks.
- Knowledge of GenAI approaches: model architectures, fine-tuning, transfer learning.
- Advanced programming skills: Python and data preprocessing (Pandas, Xarray).
- Understanding or an interest in the fields of Glycobiology, Glycomics.
- Mass Spectrometry & Experimental techniques (optional).
- Track record of published academic papers.