We are interested in the following research areas connected to the molecular biology of cancer, and more broadly, to human development and disease. We aim to develop novel computational tools, and provide biological insight using large-scale datasets from high-throughput experiments.
Biological noise in disease development
The various molecular processes of a cell are coordinated by complex gene regulatory networks. However, these processes are inherently noisy, and even isogenic cell populations show variance in RNA and protein expression levels. Increasing evidence shows that expression and splicing variance plays important roles in disease development by leading to aberrant gene regulatory network states. This variance is influenced by DNA sequence, histone modification patterns, chromatin structure, or the binding affinity of transcription and splicing factor binding sites, among other things. We are developing methods to characterize expression variance and its determinants, model the effect of variance on gene regulatory networks, and characterize it in different cancer types.
RNA-sequencing based diagnostic methods in cancer
DNA-sequencing based mutation, indel, CNV and structural variation detection is a part of routine clinical diagnostics in many cases. However, mutations can only exert an effect if they are expressed. Additionally, RNA-based sequencing methods can provide useful information besides mutation status, including allele/mutation specific expression or splicing patterns. We are working on RNA-sequencing based analysis methods to improve initial cancer diagnosis, disease subtype classification, and the selection of treatment options. We aim to develop methods robust to experimental biases and technical limitations, usable at the single patient level.
Bioinformatics and biostatistics support
We collaborate on multiple projects and provide bioinformatics and biostatistics support for analyzing targeted Illumina panel sequencing results, NanoString based experiments, histological sections or other experimental data types.