Systems Metabolomics

Understand the role of metabolism and metabolic individuality in the development and treatment of diseases using systems metabolomics.

Using metabolomic data, the main objective of our research is to identify metabolic mechanisms that translate genetic risk factors and their interplay with lifestyle and environmental factors into the development and progression of complex diseases, including Alzheimer's disease and chronic kidney disease. Thereby, we have a major focus on investigating how a person's individual metabolic make-up, its changes over time, and its link to genetic variation affect health and disease. We use metabolomics and other omics data from large epidemiological cohorts in combination with advanced computational approaches as tools to access and understand relevant metabolic individuality and its determinants in healthy populations. This forms the basis for elucidating the role of metabolic disruption in age-related diseases, their co-occurrence (multi-morbidity), and heterogeneity in a systems medicine context. Our ultimate goal is to translate our results into applied precision medicine by shifting metabolomics from a valuable research tool to a practical clinical instrument for monitoring metabolic health.

Unravel the determinants of metabolic individuality in data from large cohorts by mapping out significant associations of metabolites with intrinsic (e.g., age) and extrinsic (e.g., life style) factors.

We aim to understand the different factors influencing a person’s individual metabolome. These factors include intrinsic, non-modifiable features such as age, sex, and genetic variation as well as potentially modifiable features such as disease, medication, physical activity and nutrition. We utilize metabolomics and other omics data from big epidemiological cohorts to perform large-scale genome-wide and metabolome-wide association studies (GWAS/MWAS). Here, the metabolome can serve as an intermediate readout, connecting the genetic disposition of individuals with various diseases and risk phenotypes.

Investigate metabolic disruptions in neurodegenerative diseases.

Our computational neurobiology team is studying failures in multi-omics regulatory networks in neurological, neurodegenerative, and mental disorders. The focal omics level in our analyses is the metabolome that we use as intermediate readout for disease risk, state, stage in progression, and resilience. Using advanced computational approaches, we use this readout and interface it with genomic, transcriptomic, and proteomic markers to build multi-level frameworks that can be interrogated to identify functional hypotheses across all available molecular and regulatory layers.

Our work is embedded in several international collaborations and consortium efforts funded by the National Institutes of Health. Our current research focus is on Alzheimer’s Disease (AD), where we are partners in the Alzheimer’s Disease Metabolomics Consortium (ADMC), the Accelerating Medicines Partnership – Alzheimer’s Disease (AMP-AD) and Molecular Mechanisms of the Vascular Etiology of Alzheimer’s Disease (M2OVE-AD) consortia, and Major Depressive Disorder (MDD) within the Mood Disorders Precision Medicine consortium (MDPMC).

Provide decision support for nephrologists to personalize treatment of chronic kidney disease based on metabolic networks.

Chronic kidney disease (CKD) is a common and complex disease. It is one of the leading causes of death worldwide and is characterized by varying disease progression patterns and multiple comorbidities.

Our CKDNapp team is part of the BMBF-funded e:Med junior consortium "CKDNapp - Chronic Kidney Disease Nephrologist's App" (www.ckdn.app).  The consortium is developing a clinical decision support software (CDSS) to assist the practicing nephrologist in personalized treatment of chronic kidney disease patients. Our CKDNapp will predict adverse medical events and disease progression, refine diagnosis of CKD staging, return transparent reasoning for all predictions and recommendations, offer in silicio modification of patient parameters by the physican, and will deliver comprehensive literature support. It will be made available as an easy-to-use software for smartphones, tablets and desktop computers.

CKD atlas: In addition to CKDNapp, we are looking to integrate multi-omics data to build a resource that can help biologically justify the results provided by CKDNapp. Indeed, CKDNapp uses machine learning methods to computationally model the complex CKD system and these models do not provide biological justification for the decisions they make. Further to the biological justification, CKD atlas will offer researchers the opportunity to gain new insights into metabolic profiling and pathophysiological mechanisms related to chronic kidney disease.

Explore dynamic changes of the human metabolome in response to metabolic challenges such as diet and exercise.

The human body must continually adapt and dynamically respond to physiological challenges, such as food intake, exercise or stress. Metabolite profiles taken at multiple time points during or immediately after a specific challenge allows to monitor this systemic metabolic adaptation in a time-resolved manner, i.e., metabolomics enables us to watch metabolism ‘at work’.

On the one hand, we use challenge studies and time series metabolomics data to untangle the complexity of metabolic responses and their individuality as observed in healthy populations. A better understanding of these processes will help in optimizing and personalizing non-pharmacological disease treatments or prevention through diet and exercise.

On the other hand, we are interested in detecting features of impaired responses that could serve as early signs of a disease (i.e., analogous to oral glucose tolerance tests that allow to diagnose diabetes earlier compared to checking fasting glucose only).

In addition to time series data collected over a short period of time following a specific trigger/challenge, we use longitudinal metabolomics data with repeated measurement over several years for long-term health monitoring in large cohort studies and for following metabolic trajectories of disease.

Elucidating the complex molecular underpinnings of health and disease through integrative multi-omics networks.

Advances in high-throughput technologies have led to the generation of vast amounts of data profiling different biological layers, including DNA sequencing data (genomics), RNA expression levels (transcriptomics) and metabolite levels (metabolomics). Analysis of these single-omics have provided us with valuable insights, but fail to take the complex interplay between these layers into account. Using networks as a flexible and mathematically well-defined framework, we are working on ways to integrate, visualize and analyze heterogeneous, multi-omics datasets.

For complex diseases, which have been linked to perturbations across molecular layers (e.g., transcriptional changes, altered abundances of proteins and metabolites), this integrative analysis has the potential to guide the identification and prioritization of novel therapeutic targets and drug repositioning candidates, as well as provide further insights into the underpinnings of these heterogeneous diseases (e.g., Alzheimer’s disease and chronic kidney disease). Furthermore, we provide easy access to these integrative resources through freely accessible web services, such as the AD Atlas.

Provide easy access to metabolomics results and analysis for the wider scientific community.

With our web-services we enable the exploration and interactive visualisation of metabolomics results in our studies. Data sets with huge numbers of metabolite associations or time courses, that would be otherwise hidden in supplemental tables of publications, become thus easily available in online repositories as source and inspiration for further investigations. (OMICSCIENCE, HuMet Repository, metabolomicsGWAS Atlas, proteomicsGWAS)

Our AD Atlas and the SNiPA web-tools go one step further with the integration and annotation of knowledge from public available resources across multi-omics layers. They provide intuitive methods for data queries and show results in the context of previous findings to explore and understand underlying (patho)mechanisms.