Projects


HiGHmed

HiGHmed is a medical informatics project, which promotes the digitalization of the german healthcare system. HiGHmed is sponsored by the “Federal Ministry of Education and Reseach” and the “Medical Informatics Initative Germany”. Different locations in germany are associated with the HiGHmed project. Among them are the university hospitals of Heidelberg, Hannover, Göttingen, Berlin, Münster, Kiel/Schleswig-Holstein and other partners. Andreas Beyer is leading the HiGHmed location in Cologne.


Use Case CORD-MI (Collaboration on Rare Diseases)

In Germany, around 4 million people suffer from a “rare-disease”. Rare diseases are defined by their prevalence, which does not exceed 5 in 10.000 people, and currently there are about 8000 known rare diseases. They constitute a very heterogenous group of diseases that can severely reduce life expectancy and quality of life of affected patients. Patients often lack appropriate support and treatment, due to a small number of geographically scattered medical experts and lack of sufficient knowledge.

To improve awareness of rare diseases, 34 Medical Centers for Rare Diseases (Zentren für Seltene Erkrankungen, ZESE) were established within Germany. The major task of these centers is to improve the health conditions of patients by providing targeted diagnosis, long-term clinical care and treatment and by promoting collaborative clinical and basic research, as well as training and education.

In CORD-MI, the Medical Informatic Initiative works together with ZESE, ACHSE and NAMSE with the goal to facilitate the transfer of knowledge between medical centers and to ensure a concerted action on the national level to guarantee adequate treatment and care for each patient. Prof. Andreas Beyer and Prof. Jörg Dötsch (Director of ZESE , University Hospital Köln) are the leaders of the CORD-MI project at the location in Cologne.


CODEX

The Network University Medicine (NUM) is building the CODEX research data platform, a secure, extensible and interoperable platform for providing COVID-19 research data that connects university hospitals nationwide. This data protection-compliant infrastructure will be able to map complex COVID-19 research data sets, including clinical data, image data and biospecimen data, in a multicentre, patient-related and pseudonymised manner and make them available to research.
The MeDIC creates the organisational and technical framework for the provision of data from the primary clinical systems and also provides the technical framework for the integration of new software components and processes at the UKK.


B-FAST

The aim is to develop transferable surveillance and testing strategies and to test them in different areas of application. In addition to random samples from the population, data from risk areas such as hospitals, retirement homes, schools and day-care centres as well as cultural and sports facilities are included in the study. B-FAST has the primary goal of developing a sustainable, scalable surveillance and testing strategy for future pandemics.

The MeDIC is responsible for implementing the central component SmICS (Smart Infection Control and Surveillance (https://highmed.org/news/highmed-smart-infection-control-system-version-1-for-covid-19-available)) at UKK.


Predicting the constrained evolution of tumors

We are part of the “Collaborative Research Center 1310 Predictability in Evolution”.
The mutational pattern of a tumor – i.e., the set of mutations observable in a given tumor – is shaped by physical and functional constraints acting during its evolution. This project aims to predict the occurrence and selection of specific mutations as a function of molecular and cellular co-factors, which will provide new insight into the mechanisms of tumor evolution. We propose to systematically characterize molecular and cellular factors (1) acting on baseline mutation rates and (2) driving the subsequent selection of mutations during tumor evolution. Using whole-genome and exome sequencing data, copy number data, and RNA sequencing data for hundreds of tumors we will create models predicting mutations as a function of genomic position and tissue type (plus other covariates, e.g. epigenetic data). Second, we will use functional information, as well as network- and pathway data, to model fitness effects of mutations during tumor evolution. In combination, these two models will predict in vivo frequencies of tumor mutations. The long-term goal of this project is to predict the most likely mutational pattern of a tumor given the tissue type (cell of origin), expression of other genes, and the patient’s history.


Execution of Cell Fate Choice by Cooperative Gene Activities

This is a joint project together with the group of Dr. Martin Leeb (Max Perutz Labs in Vienna) that is jointly funded by the German DFG and the Austrian FWF.

Research in recent years has contributed to a substantial understanding of the molecular underpinnings of self-renewal and pluripotency. However, the factors that compute cues from the cellular environment to elicit an integrated and exact series of cell-fate choices remain largely elusive. Their identification and knowledge of their molecular role will be a crucial milestone in devising efficient and deterministic protocols to drive lineage specification. To this end, we will combine genetic screens, detailed gene- and pathway-level analysis and systems biology approaches to explore the mechanistic basis of the key cell fate decision that defines the exit from pluripotency. We have established haploid ES cells as a powerful screening platform to interrogate developmental processes and have performed a saturation screen to identify central players involved in the exit from self-renewal. We propose to use high-throughput cell culture approaches combined with sophisticated and customized computational tools to allow us to approach ES cell differentiation in a systems biology approach. Specifically, one main aim of this project is to delineate the genetic switches that gate the exit from naïve pluripotency by studying a series of genetic perturbation of differentiation. The combination of expertise in the Beyer and Leeb laboratories will position us ideally to identify relevant candidate genes, regulatory nodes and to delineate the exact molecular mode of action of candidate-genes and pathways using controlled genetic and biochemical experiments under defined culture conditions.


Gene-regulatory networks in healthy and diseased podocytes

We are conducting this project together with Dr. Martin Kann within the KFO 329 “Disease pathways in podocyte injury – from molecular mechanisms to individualized treatment options”.
In response to damage, podocytes employ mechanisms involving the cell nuclei to change gene expression patterns by activating Tgf-beta, Notch and Wnt signaling as previously shown in hu-man biopsy specimen as well as in mouse models of FSGS. Longstanding knowledge of human and murine hereditary FSGS established mutations in nuclear proteins such as WT1 as disease causing. However, such nuclear mechanisms and gene regulatory networks in podocytes are not well understood. In this context, the transcription factor (TF) WT1 plays an important role as it is indispensable to podocyte development and maintenance.

Here, we aim to test (1) the hypothesis that in healthy podocytes, WT1 orchestrates the binding of further podocyte TFs, which then assume specific functions in podocyte physiology by controlling podocyte gene regulatory networks. To this end, we are using ChIP-seq on relevant transcription factors as well as RNA-seq on FACS-sorted podocytes. (2) We are modulating podocyte gene regulatory networks by induction of podocyte disease and assay transcription factors downstream of signaling pathways that are activated upon injury. Specifically, we directly damage mouse podocytes by adriamycin and deplete their numbers using a diphtheria-toxin model, both resulting in an FSGS phenotype. Podocyte gene regulatory networks will then be assessed by ChIP-seq for effectors of Notch-, Tgf-beta-, and Wnt signaling pathways. (3) We will translate key podocyte gene regulatory network nodes into human context using biopsy specimen from patients with FSGS and novel mouse models to further elucidate the intracellular signaling pathways that trigger podocyte dysfunction.


Mechanisms of age-associated decline in splicing fidelity

This project is funded by the Center for Molecular Medicine Cologne (CMMC).

Recent work by us and others has demonstrated that the age-associated decline of cellular functions is associated with the decline of RNA splicing fidelity. We have shown that during ageing the speed of RNA Pol-II elongation increases, which leads to increased errors during transcription and splicing. Further, we could show that slowing down transcription improves the quality of transcripts and extends lifespan in different metazoan species. However, we lack understanding about the molecular factors contributing to this functional decline. In this project we aim to (1.) identify tissue-specific and tissue-independent splicing ‘mistakes’ at high single-cell resolution, (2.) identify molecular co-factors improving or worsening splicing fidelity, and (3.) thus better understand effects of splicing fidelity on cellular fitness. We propose to tackle the above questions by performing extensive single-cell profiling in order to associate the emergence of specific splicing events with other molecular co-factors, such as the expression of splicing regulators, Pol-II components and chromatin regulators. These experiments will be conducted in livers and kidneys from mice of different age and lifespan-extending interventions (dietary restriction). Importantly, we will employ innovative technologies that will allow us to profile chromatin states and RNA levels in the same cells, which will enable the direct correlation of chromatin state changes with splicing and gene expression changes. This project will heavily rely on single-cell bioinformatic methods developed in the Beyer lab.