Fig. 1

This boxplot figure presents the results of the machine learning predictions. We created three different datasets, one which includes all patients less than 90 years old, the second includes every patient less than 80 years old, and the third with every patient less than 70 years old. These are indicated as the oldest birthyear “data.” Each dataset included an equal number of patients with a “severe reaction” to COVID-19 and an equal number of age-matched people drawn from the general UK Biobank population, “normal.” For comparison, we took those three datasets and randomly permuted the status (“severe reaction” or “normal”) and repeated the process. This randomly permuted dataset is labeled oldest birthyear “random.” For each dataset, we repeated the whole process 100 times, each time with a different set of age-matched people from the general UK Biobank population