Ewa Szczurek

Research Interests

  • Computational Biology, focus: Computational Oncology
  • Statistical data analysis and machine learning

Current projects

  • Modeling epidemic spread. I am involved in two projects modeling COVID-19 dynamics in Poland: SEIR-based COVID-19 model and microsimulation-based MOCOS (MOdeling COronavirus spread) model.
  • Modeling tumor microenvironment What is the spatial organization of the tumor and its neighborhood? How do they interact? These interesting questions are solved in our lab using spatial transcriptomics and tumor imaging data
    Types of methods to be developed in this project: probabilistic graphical models, machine learning models
  • Drawing the genealogical trees of tumors. Which cancer mutations come first? How do metastases occur? How does drug resistance appear in cancer? Is tumor evloution neutral, or is it driven by selection? These and many more questions about the family history of tumor cells I find very exciting! Find out more on our CONTRA (Computational ONcology TRaining Alliance) website.
    Types of methods to be developed in this project: probabilistic graphical models, mathematical models
  • Working out the bottleneck of metastasis initiation. Cancer makes astronomic numbers of trials to set up metastatic colonies. Only a few succeed. We figure out why.
    Methods developed: mathematical models
  • Who talks with whom in cellular networks? I study cellular signaling pathways by kicking their members out and observing what happens. And of course, we model cancer pathways!
    Methods developed: probabilistic graphical models
  • Fighting cancer with its own weapons. I am analyzing the phenomenon of synthetic lethality between cancer genes. We can use synthetic lethality for cancer treatment. Once one gene is mutated, we can target its synthetic lethal target in order to kill cancer cells. Learn more on our project website.
    Methods developed: statistical tests, survival analysis methods
  • Deep Pathologist! Can deep learning algorithms scale the work of pathologists? We train a deep pathologist on colorectal cancer histological images to recognize eight tissue types.
    Methods developed: deep learning models
  • Modeling efficacy of drugs and drug combinations. Together with colleagues from the Oncology Bioinformatics department at Merck, Germany, we try to understand and predict how drugs work on cancer cell lines.
    Methods developed: statistical models, optimization algorithms


Winter semester
  • Statistical data analysis 2 (in English), materials available for students at Moodle
Summer semester
  • Statistical data analysis 1, materials available for students at Moodle (in Polish)
Previous courses

Organized seminars

Past organized conferences