Whole-genome epigenetic and multimodal liquid profiling for translational biomarker discovery

• Whole-genome epigenetic readout. cfMeDIP-seq delivers genome-wide methylation profiles from low-input plasma or CSF, witho​ut bisulfite conversion. As a single, transferable assay, it provides a common language across very different disease contexts.
• From plasma into CSF. In neuropsychiatry we move beyond plasma and apply cfMeDIP-seq to cerebrospinal fluid, where the brain-derived methylation signal is far less diluted and cell-type-specific patterns are more readily resolved.
  
• Multi-layered liquid biopsy. Methylation (cfMeDIP-seq) is paired with cell-free RNA (RAREseq) and with platelet-associated cfDNA, opening complementary readouts on the same blood draw — useful where DNA shedding alone is limiting.
• Anchored in deeply phenotyped cohorts. The LIFE-Adult cohort and the Leipzig Medical Biobank allow plasma methylomes and​ protein biomarkers to be paired with detailed clinical, endocrinological and psychiatric phenotyping at sample sizes uncommon in early-stage liquid-biopsy research.​
  

Three connected fields, listed in order of current emphasis:

1. Oncology

cfMeDIP-seq based methylation signatures complement mutation-centric ctDNA assays: they are independent of any single actionable variant and they carry tissue-of-origin information. We pursue this for early detection of molecular relapse and minimal residual disease (MRD) after curative-intent treatment, and for tumor-fraction estimation in plasma where mutations are below detection. Comprehensive genomic profiling is delivered jointly with the Institute of Human Genetics, Leipzig.

2. Neuropsychiatric disorders

Psychiatric and neurodegenerative diagnoses still rest predominantly on clinical assessment. We use cfMeDIP-seq on cerebrospinal fluid to read methylation signatures of neuronal, astrocy​tic, oligodendroglial and microglial origin, and to ask whether they track disease activity in schizophrenia, depression and dementia, and help separate primary psychiatric disease from neurodegenerative differentials. On the genetic side, the group contributes to the international GRIN2A registry, characterising the spectrum and treatment response of GRIN2A-related neuropsychiatric phenotypes (Lemke et al. 2025).

3. Metabolism

At the interface of endocrinology, mood and metabolic disease the group has a long-standing focus on the stress-induced cytokine GDF-15 (and its receptor GFRAL) and on ghrelin signalling. Within the LIFE-Adult cohort we examine how circulating GDF-15 relates to depressive symptomato​logy across the lifespan (de Vos et al. 2026) and to hepatic steatosis and fibrosis (Dietzel et al. 2026). These projects complement the epigenetic-liquid-biopsy work by anchoring it in measurable endocrine and metabolic physiology.

Core: cfMeDIP-seq

Cell-free methylated DNA immunoprecipitation sequencing enriches methylated cfDNA fragments using a 5-methylcytosine antibody, delivering genome-wide methylation profiles from low-input plasma or CSF without bisulfite conversion and at modest sequencing depth. It is the technical backbone for our oncology, neuropsychiatry and metabolism projects.​​

Extensions in implementation

• Cell-free RNA via RAREseq. Ultrasensitive sequencing of low-abundance cell-free RNA species, adding a transcriptional layer to the methylation readout — particularly relevant for tissue-of-origin and disease-state inference where DNA shedding is sparse.
• Platelet-associated cell-free DNA. Workflows that recover platelet-bound DNA fragments and integrate them with plasma cfDNA, with the goal of improving sensitivity for low-shedding tumors and of openin​g additional methylation-based biomarker channels.
  
  

Computational analysis

Genome-wide methylation, sequencing and multi-modal biomarker work generates data volumes and analytical demands that are challenging. We rely on modern high-performance computing infrastructure and on dedicated bioinformatics pipelines for cfMeDIP-seq and cell-free RNA processing. Beyond standard preprocessing, our analytical work makes systematic use of machine-learning approaches — supervised classifiers, methylation-based cell-type and tissue-of-origin deconvolution, and signature discovery — applied to large reference and discovery cohorts. A particular focus is the joint analysis of multi-omic data (methylome, transcriptome, protein and deep clinical phenotyping) from the same individuals, in order to identify biomarker signals that are robust across modalities.​

Supporting techniques

• ddPCR and qPCR for highly sensitive single- and low-plex quantification of established hotspot mutations.

Quantitative immunoassays (ECLIA / ELISA) for GDF-15, ghrelin and related analytes.

Clinical specimens are sourced via the Leipzig Medical Biobank and through prospective recruitment with partner clinics at the University of Leipzig Medical Center. Population-based work us​es the LIFE-Adult cohort (Leipzig Research Centre for Civilization Diseases). Comprehensive genomic profiling is run jointly with the Institute of Human Genetics, Leipzig.

Dietzel NA, Schmidt M, Wiegand J, Berg T, Biemann R, Baber R, Kluge M, Wirkner K, Wittekind DA. Growth Differ​entiation Factor 15 (GDF-15) as a modulator of hepatic steatosis and fibrosis: insights from a 6-year retrospective cohort study. Frontiers in Medicine 2026;13:1733339. doi:10.3389/fmed.2026.1733339.

de Vos C, Mergl R, Biemann R, Baber R, Riedel-Heller S, Wirkner K, Kluge M, Wittekind DA. Age-dependent associations between growth differentiation factor 15 (GDF-15) and depressive symptoms: evidence from a population-based cohort. Journal of Affective Disorders 2026;407:121832. doi:10.1016/j.jad.2026.121832.

Lemke JR, Eoli A, Krey I, Popp B, Strehlow V, Wittekind DA, et al. GRIN2A null variants confer a high risk for early-onset schizophrenia and other mental disorders and potentially enable precision therapy. Molecular Psychiatry 2025;31(1):374–382. doi:10.1038/s41380-025-03279-4.

Wittekind DA, Kratzsch J, Mergl R, Wirkner K, Baber R, Sander C, Witte AV, Villringer A, Kluge M. Childhood sexual abuse is associated with higher total ghrelin serum levels in adulthood: results from a large, population-based study. Translational Psychiatry 2023;13(1):219. doi:10.1038/s41398-023-02517-z.

Wittekind DA, Scholz M, Kratzsch J, Löffler M, Horn K, Kirsten H, Witte V, Villringer A, Kluge M. Genome-wide association and transcriptome analysis suggests total serum ghrelin to be linked with GFRAL. European Journal of Endocrinology 2021;184(6):847–856. doi:10.1530/EJE-20-1220.