Multi-omic analysis of post-acute sequelae of COVID-19 (PASC)

Millions of patients surviving COVID-19 infection develop post-acute sequelae of COVID-19 (PASC), a debilitating condition that severely disrupts quality of life. Limited data exists on PASC patient peripheral blood multi-omic landscape. We enrolled 155 individuals with self-reported PASC who were compared with 100 acutely ill COVID-19 patients and with 44 healthy subjects. State-of-the-art mass spectrometry and next-generation leukocyte RNA sequencing were conducted in the blood samples of these participants. PASC status was associated with failure to regain a healthy baseline molecular status in these patients and aberrations in multiple biological pathways, including sphingolipids, chromatin-associated processes, and others. A machine learning algorithm trained on patients’ data produced 9 top entities allowing PASC diagnosis with a mean precision-recall area under the curve (AUC) of 0.97. An interface tool was generated to facilitate rapid access to the presented data and enable further analyses. This work provides new mechanistic insights into PASC through integrated multi-omic and machine learning analyses, establishing a robust exploratory framework and data resources to guide future biomarker validation and clinical meta-analyses.