In neurodegeneration, epigenetic modifications are emerging as stable, disease-relevant signatures that appear early in disease progression. cfDNA, released during cellular turnover and injury could act as a potential biomarker capable of detection in the preclinical phase. However, only about 2% of total cfDNA in circulation is neuron-derived. The low abundance makes detection difficult, contributing to why this area remains underexplored and yet to be integrated into clinical screening for neurodegeneration. Hence, we aim to develop a blood-based cfDNA methylation signature panel (identified using machine learning) using sensitive technology like nanopore. In addition, we want to improve this panel’s specificity and sensitivity by incorporating other omic markers into the panel for neurodegeneration.
Objectives:
* Identify neuron-specific cfDNA methylation markers, along with other omic signatures using machine learning analysis of large-scale public methylome datasets.
* Validate shortlisted markers with Oxford nanopore sequencing for high-resolution methylation profiling.
* Translate validated markers into a clinically implementable assay
Rationale:
The absence of reliable, minimally invasive diagnostics for neurodegeneration represents a critical barrier in clinical neuroscience. While imaging and CSF-based tests provide diagnostic information, their invasiveness, cost and late-stage applicability limit their impact. Protein markers are typically detectable only at later stages of the disease, limiting their utility for early intervention.
Omic signatures offers a transformative alternative, with the ability to detect neuron-derived signals from peripheral blood. Importantly, cfDNA methylation profiling would enable determination of cell-type origins, surpassing protein biomarkers in specificity. Established cfDNA-based assays in oncology, prenatal care, and transplantation, demonstrates their reliability as non-invasive diagnostics.
