Identification of a molecular signature associated with covid-19 severity using a comprehensive 1H-NMR serum metabolomics profiling strategy

Background and Aims: 

Atherogenic dyslipidemia together with specific pro-inflammatory profiles have been identified as markers of severe adverse outcomes in COVID-19 disease. This study aims to evaluate nuclear magnetic resonance (NMR)-based metabolomics combined with machine learning algorithms as a strategy to identify a molecular signature associated with poor CoVID-19 prognosis.