Autoantibody Profiles in COVID-19 Cohorts

Statistical details of experiments can be found in the figure legends. All REAP screens, ELISA, and neutralization assays were performed with two technical replicates. Data analysis was performed using R, Python, Excel, and GraphPad Prism. No statistical method was used to predetermine the sample size. Patients were excluded from the acute COVID-19 cohort if they did not have moderate or severe disease or if they only had a 1-time point because this would prohibit the longitudinal analysis required by our study. Patients were excluded from the vaccination cohort if an unclear sample label or obvious cross-well contamination was detected. Additionally, one patient was excluded due to failed REAP process and lack of autoantibody data. Two patients were excluded from the Myocarditis cohort: 1 patient received IVIG before their blood sample, which complicates the results of REAP, making them uninterpretable; 1 patient had onset of myocarditis 21 days after the vaccination, which is an atypical time course and thus the cause of the myocarditis (viral vs. vaccine) could not be determined. The experiments were not randomized.
For the acute COVID-19 cohort and the Benaroya cohort, REAP and ELISA/neutralization studies were performed by the investigator before receiving associated clinical annotations. For the Yale HCW cohort, CoronaVac cohort, and the myocarditis cohort, the investigator received clinical annotations at the time of REAP/ELISA/neutralization studies. However, samples were analyzed in the same manner in a randomized plate layout.