Biomarker discovery for disease, progression and treatment by metabolomics

project home: Greifswald

State of the art

Workplan The project consists of five parts with a high degree of interrelation integrating patient stratification and biomarker evaluation with basic research and a translational component to the general population. Notably, every work package (WP) is designed to be substantial even in the case one of the others will not yield the expected results.
WP1 is dedicated to extensive pre-processing of the supplied metabolomics data, including missing value imputation, outlier detection and derivation of valid quality measures. To complement metabolome-wise analysis in WP2/WP3 a data-driven metabolite network will be compiled. Key objectives of WP2 are the identification of subgroups of patients and identification of biomarkers to characterize these groups. To this end, linear regression analyses will be used to screen for disorder/symptom–metabolite associations which will be subsequently aggregated using pathway enrichment tools. Patient clustering will be done based on molecular profiles both in a supervised and unsupervised manner establishing a correspondence between current clinical practice and novel molecular profiling. Candidate metabolites will be further integrated with phenotypic information using network techniques based on Random forests to compile a metabolite-disease network enabling systemic inter-pretation of mitochondrial disorders.WP3 is dedicated to the repeated measures for about one third of the patient population, deriving biomarkers for disease progression and evaluating those in follow-up data. Further, subgroups and biomarkers identified in WP2 will be followed up to assess their stability and hence reliance for clinical purposes. Search for interactions in time between metabolic profiles and disease progression will provide insights in the metabolic flexibility of patients.
Key objective of WP4 is the metabolic characterization of sequence variants for the patients to map orphan mitochondrial genes to metabolic pathways. To this end, we will leverage our vast data on gene-metabolite associations in healthy populations.   
WP5 aims in the translation of genetic findings from WP4 to the general population. I.e. performing mitochondrial-genome-wide association studies on candidate metabolites obtained in WP2 and WP3 using the Study of Health in Pomerania (SHIP) and other studies in the CHARGEmtDNA+ working group. Refinement of biomarker signatures from WP2/3 will also be based on data from healthy subjects in the deeply characterized SHIP. As an example, SHIP will be used to test for common confounders possibly hampering implementation of the biomarkers into clinical practice.


Leveraging metabolomics data from patient with repeated samples taken at multiple time points during the course of the disease, and integrating genetic information from these patients we here aim:
1. to improve our understanding of the pathobiology of mitochondrial diseases with heterogeneous phenotypes by highlighting those metabolic pathways perturbed across all mitochondrial disease and those altered with specific mutations,
2. to identify metabolite biomarkers for mitochondrial diseases and their prognosis,
3. to assign potential metabolic activities for orphan mitochondrial genes, and
4. to reveal potential new therapeutic strategies aimed at complementing those metabolic pathways suppressed in mitochondrial disease and eliminating potential toxicants that accumulate.

Project management and contact address

Maik Pietzner, Dr.

Universitätsmedizin Greifswald

Ferdinand-Sauerbruch-Straße, 17475 Greifswald

Project members

Gabi Kastenmüller, Dr.

Helmholtz Zentrum

Ingolstädter Landstraße 1, 85764 Neuherberg

Phone: +49 89 - 3187 - 3578

Fax: +49 89 - 3187 - 3585