Computational pathology and transcriptomics

We use deep learning for exploring whole slide images and combine the extracted features with molecular data for gaining new insights into the cancer universe. We were the first to demonstrate that molecular subtypes of colon cancer can be predicted using the pathology images.

Microbiome and cancer

Linking the microbiome to tumor progression.
The microbiome modulates the host immune response to cancer and plays a central role in the tumor progression. Understanding these interactions may provide prognostic and predictive tools and hints towards novel therapeutical approaches.

Tumor heterogeneity

Bioinformatics explorations of intra-/inter-tumor heterogeneity.
Both inter-tumor and intra-tumor heterogeneity are major obstacles towards precision medicine. We develop bioinformatics approaches combining various OMICS for shedding a light into this complex universe.