Single-cell sequencing technologies (e.g. scRNA-seq) have transformed personalized medicine by providing high-resolution biomarkers for diagnostics, survival analysis and treatment response prediction. However, applicability of scRNA-seq technology in statistical treatment response analysis is limited due to special characteristics of the scRNA-seq data. In this project we will develop probabilistic deep learning methods for these tasks. We will extensively evaluate and validate our methods on various single-cell datasets including datasets from cancer patients treated with latest immunotherapies.

2026

2029