Birkir Reynisson

PhD project by Birkir Reynisson

Name: Birkir Reynisson
Project Title:
T Cell Immunoinformatics
Immunoinformatics and Machine Learning
Morten Nielsen and Paolo Marcatili

Project Description:
The main objective of this project is to develop sequence based models of a receptor interaction that lies at the heart of immune regulation: the interaction between Major Histocompatability Complex(MHC), peptides and T-Cell Recptors (TCR). Cells signal their interactions and internal state via peptides bound to the cell surface receptor MHC. Peptide-MHC complexes are scrutinized by T-cells and stable interaction between pMHC and TCR can initiate an immune response. Peptides that can elicit an immune response are called epitopes. MHC and TCR protein sequences vary between individuals leading to diverse peptide binding repertoires and immune recognition. In my PhD I build neural network models that learn patterns linking MHC sequence to antigen presentation repertoires.

The immune system is our bodies line of defence against pathogens and cancer. Understanding the mechanisms that underlie it´s regulation has great clinical applications. In rational vaccine design, pathogen or cancer epitopes are identified in silico with sequence based models, such as those developed in this project. Protein drug deimmunization is the process of removing epitopes from protein drugs by mutations, guided by prediction. These are worthy goals and their outcome can only be as good as the models they rely on.




Morten Nielsen
Gruppeleder, Professor
DTU Sundhedsteknologi
45 25 24 25


Paolo Marcatili
DTU Compute