Primary immunodeficiencies are rare diseases caused by defects of the immune system. They can
lead to severe and recurrent infections, malignancies and autoimmunity, but are very hard to
diagnose. As there exist many subtypes, it is crucial to identify correctly which parts of the immune
system are affected to enable optimal treatment.
One way to gain insight in the status of the patient’s immune system is by using a technique called
“flow cytometry”. Given a patient sample (e.g. a blood sample), this technique measures a certain
set of surface proteins on all of the individual cells, resulting in a detailed overview of which cell
types are present and in which ratios.
In the last decade, the number of proteins measured simultaneously has increased from about six
to more than thirty. While this detailed information is extremely valuable, the amount of data
collected also becomes harder to handle. Computational tools can help, but the current solutions
are still insufficient to tackle all challenges, such as detecting very small proportions of clinically
relevant cell populations between all other cells.
My goal with this project is to improve these computational methods, taking into account the
specific setting of primary immunodeficiencies. Overall, these algorithms will allow immunologists
to better exploit the high-dimensional cytometry data, enabling them to study the immune system
in a more systematic and unbiased way.