Machine learning in personalised medicine
Although the image may imply otherwise, personalised treatment choices in cancer and other diseases are not as futuristic as it seems. A very large amount of genetic data in combination with clinical data in now readily available. It has the potential to substantially support decision making in medicine. However, such data have become incredibly complex, so that a physician alone would struggle to make use of them. The key to bring the physician together with the data is to design biology informed, transparent and interpretable machine learning algorithms that can aid trustworthy clinical decision making.
We contribute to multiple projects and consortia to make it a reality to increase safety of genome editing, aid cancer diagnosis, and overcome the COVID19 pandemic with using machine learning strategies as part of the DeCOI consortium.
Specifically for cancer treatment, we are part of the HEROES-AYA consortium to study tumour heterogeneity, evolution, and resistance on fusion-gene-driven sarcomas. More about the project (for now only in German) can be found here. A link to DeepL to translate it, can be found here. For this project we are looking for a postdoc.