Fig. 2: Nasal cell lineage abundance during and after COVID-19.
a, Stacked bar charts showing the composition of the nasal immune system in acute COVID-19 (red), during ERS (after ICU but still in the hospital; orange) or in COVID-19 convalescence (5–6 weeks after hospital discharge; pink) or in healthy donors (HD; blue). b, Correlation analyses between nasal immune and epithelial cells for hospitalized individuals (left) and convalescent individuals and healthy donors (right) are shown. Ranks of individuals are shown with color corresponding to group as well as a regression line (blue), 95% confidence intervals (shaded area) and results from Spearman correlation analysis. c, Ratio of nasal immune cell types normalized to the number of epithelial cells from the same sample. Individuals and box plots are shown, and paired samples between are indicated by gray lines. If a cell type was not detected in at least one sample, half the value of the lowest recorded number was added to each sample before log transformation. d, Heat map showing log10 relative abundance (RA) of nasal cell clusters scaled per cluster. e, Box plots of nasal clusters during acute COVID-19. Box plots depict median and interquartile ranges, with whiskers extending to 1.5× interquartile range or maximum value; *P < 0.05 and ***P < 0.001 by linear mixed model with group as fixed effect and individuals as random effect with post hoc testing and Tukey multiple testing correction, followed by Benjamini–Hochberg correction for comparing multiple lineages or subsets. See Supplementary Tables 3 and 4 for exact test results. Samples from acute individuals (n = 9 individuals, 10 independent samples), ERS individuals (n = 11 individuals, 18 independent samples), convalescent individuals (n = 16 individuals) and healthy donors (n = 12 individuals) are used; Neutro, neutrophils. Only the first sample per donor in a time point (acute or ERS) is shown for a, c and e, but all are included in statistical modeling.
