Click here to stream the public defence
Due to copyright issues, an electronic copy of the thesis must be ordered from the faculty. For the faculty to have time to process the order, the order must be received by the faculty at the latest 2 days before the public defence. Orders received later than 2 days before the defence will not be processed. After the public defence, please address any inquiries regarding the thesis to the candidate.
Trial Lecture – time and place
See Trial Lecture.
Adjudication committee
- First opponent: Professor Frederick A. Matsen IV, Fred Hutchinson Cancer Center, United States
- Second opponent: Senior Researcher Albert Bondt, Utrecht University, Netherlands
- Third member and chair of the evaluation committee: Professor Anne Spurkland, University of Oslo
Chair of the Defence
Professor Frode Lars Jahnsen, University of Oslo
Principal Supervisor
Associate Professor Victor Greiff, University of Oslo
Summary
The adaptive immune system is capable of recognizing and combat a vast array of pathogens, including viruses, bacteria, and cancerous cells, using B- and T-cell receptors, together known as adaptive immune receptors (AIRs). Recent advances in sequencing technologies allow us to analyse this diverse repertoire of AIRs, revealing its complexity.
To understand this complexity, in particular to analyze the AIR repertoire data, the computational methods are required, particularly machine learning (ML). Such ML algorithms can be used to predict the targets of each individual immune receptor, which may help to design therapeutic drugs. Alternatively, ML algorithms can be applied to immune repertoires to diagnose past and ongoing diseases based on a single blood sample.
While ML shows promising results in predicting AIR specificity and classifying immune repertoires based on their disease status, its accuracy is limited by the lack of large, well-characterized datasets. Furthermore, studying antibodies — a secreted form of B-cell receptors which are the actual proteins responsible for immune responses, poses unique challenges. In particular, the antibody sequencing method used to analyze antibody molecules is still not thoroughly benchmarked.
In this thesis, Maria Chernigovskaya addresses these limitations by developing a simulation framework which can be used to support the development and benchmarking of ML methods for studying AIR repertoire data. Furthermore, Mariia and colleagues created a large-scale dataset for benchmarking antibody sequencing, and revealed strength, pitfalls, and limitations of this method.
The results of this work will enable researchers to develop more reliable tools for exploring the adaptive immune system at both genomic and proteomic levels in the future.
Additional information
Contact the research support staff.