PhD position in LLMs & complex reasoning for supporting scientific discovery in cancer (University of Manchester)

We have an exciting new PhD position at the interface between LLMs, complex reasoning and cancer discovery.

Melanoma is an aggressive type of skin cancer associated with an incidence of ≈17500 patients and 2300 deaths per year in the United Kingdom. Immunotherapy (checkpoint inhibitors [CPI]; anti-CTLA-4, anti-PD-1, anti-LAG3) has revolutionised patient outcomes in melanoma and other cancers however, challenges remain. Critically, the presence of liver metastasis is associated with poor CPI response and prognosis, with patients with liver metastasis progressing on average 15 months earlier than those without.

The aims of this exciting project will be to understand complex interactions between melanoma and cells within the liver microenvironment to understand mechanisms associated with immune suppression and resistance to CPI. To do this, the student will systematically integrate curated databases (e.g. KEGG, Reactome, STRING, and HMDB) alongside systematic extraction from the literature to derive models of signalling pathways, metabolic reactions, and protein-protein interactions relevant to this immunosuppressive environment. The student will develop techniques in open information extraction, integration of large datasets from multiple sources, large language models, causal reasoning, natural language inference and explainable question answering. Furthermore, they will gain an in-depth understanding of cancer biology and the immune system.

We are looking for a hard-working, focused, ambitious person to join our excellent, friendly, inclusive and collaborative teams. The student will enjoy integrating with both the Freitas laboratory which focuses on development of AI methods to support abstract, explainable and flexible inference and the Lee laboratory which uses a broad range of in vitro and in vivo techniques to study melanoma, with the aim of developing novel therapies for patients. We would be particularly happy to receive applications from individuals with a strong academic track record and Masters-level and/or other research experience in artificial intelligence, large language models and bioinformatics.

The project will provide comprehensive training in cancer biology, large language models and use of artificial intelligence to inform mechanisms behind complex tumour-microenvironment interactions and to build complex models of these.

Essential requirements:

-

BSc and MSc in computer science, bioinformatics or related areas. -

Experience in the construction of NLP and ML models evidenced by academic or industrial projects. -

Python programming. -

Confident in the development of complex computational pipelines. -

Experience in biomedical problems is a plus.

Interested applicants please send an email to andre.freitas@manchester.ac.uk and rebecca.lee-3@manchester.ac.uk with your CV by December 31st.

Additional information.

https://www.findaphd.com/phds/project/bicentenary-modelling-melanoma-induced...