Machine learning of epitopes

Royal Institute of Technology

SciLifeLab in Solna

IT Teknik Medicin Natur

Our immune system uses antibodies to identify and neutralize foreign objects such as bacteria and viruses. These potent molecules have been tailored by the immune system to differentiate between foreign and self allowing an efficient binding to parts of foreign proteins without risk of harming the body. We call such exposed parts of proteins epitopes. The actual exact process of how epitopes are selected, and what makes an epitope a good target region, is currently unknown. However, if the process was understood, there would be fargoing consequences for e.g. vaccine production and understanding of autoimmune disease.

At the Science For Life Laboratory in Stockholm we currently have assembled a large set of well-characterized epitopes, a set that makes a good training set for machine learning.

We are looking for a diploma worker that is well familiarized with modern machine learning. Depending on your interest, you could either use kernel-based techniques or deep-learning technology to learn which part of a foreign body makes a good epitope.



Programming experience from both compiled and scripting languages, and a strong interest in machine learning. Knowledge or experience from biological or medical applications are meriting but not required.


Notera att sista ansökningsdag passerats