Abstract
A key aspect to realize personalized medicine approaches is to build computational models that can simulate variation and perturbation of cellular behavior. Current large datasets of epigenome, transcriptome and drug perturbation measurements of cells allow to create machine learning models that can serve as digital twins for cellular behavior. I will present two approaches in this direction. First, digital twin models that allow the prediction of gene expression from epigenome data, which can be leveraged to conduct a new type of epigenome-wide association study in disease. Second, a model that can prioritize matching drug monotherapies from differential single cell expression data. I will illustrate how the combination of these models allows novel applications and addresses current limitations of existing methods.
Meeting the speaker
If you want to meet with Prof. Schulz, you can select a slot at https://form.jotform.com/261270876361359 or contact Anthony Mathelier at anthoma@uio.no.