The immune system can patrol cell status by recognising epitopes bound to Human Leukocyte Antigen class I (HLA-I) molecules and triggering cytotoxic T cell (CTL) activation. Epitopes arrive at the cell surface through several steps of the antigen processing and presentation pathway (APP). Despite major discoveries in the past decades, the precise dynamics of APP are still not fully understood. These dynamics, however, determine the types and quantities of epitopes and therefore regulate the CTL response.
The objective of my project is the development of an APP model from gene transcription to recognition by CTL clones that captures the underlying APP mechanisms and dynamics. Current approaches are mainly machine learning-based and focus only on parts of the APP. They neglect the underlying APP dynamics and do not elucidate novel mechanisms. Therefore, they fail to predict epitopes under altered cellular states such as infection. My project overcomes this pitfall by complementing large scale machine learning approaches with mechanistic modelling to comprehensively understand the APP and enable successful epitope prediction.
To achieve this goal, I will:
1) develop an APP integrated computational description that combines a mechanistic model of the APP (based on detailed quantitative in vitro data) with a large-scale machine learning-based multi-omics model;
2) apply the model in two therapeutic-relevant cellular systems – Measles virus infection and chemotherapy-induced senescence – resulting in the educated identification of canonical and non-canonical epitopes.
This project is at the cutting edge of the research with direct application in epitope discovery for therapies such as anti-cancer immunotherapy, vaccination against infection and modulatory immunotherapies against autoimmune diseases. The implemented computational framework will be made available to the community thereby providing a direct benefit to European scientific public and private sectors.
Fields of science
Funding SchemeERC-STG - Starting Grant