Elastic-Net Penalized Methylation-Based Classification Of Osteoarthritis In The Multicenter Osteoarthritis Study

Purpose: Currently, there are few biomarkers available to assess OA disease status. This may be attributed to the highly heterogenous nature of OA and possible underlying OA endotypes that may have differing disease etiologies. For example, pathogenesis of knee OA may be determined largely by mechanopathology while hand OA may be more highly influenced by systemic factors. Involvement of OA at multiple joint sites may encompass a distinct entity from single-site OA that results from systemic initiation rather than joint-specific biomechanical factors. Identification of biomarkers that can be used for early identification and treatment of OA is needed. Epigenetic mechanisms like DNA methylation have been shown to play an important role in OA pathophysiology and may provide novel insights into OA biomarkers. Thus, we aimed to build methylation-based joint-specific and multi-joint OA case classifiers using machine learning.