Skip to main content
. 2023 Jan 20;120(4):e2217902120. doi: 10.1073/pnas.2217902120

Fig. 4.

Fig. 4.

Broad changes to cellular networks in COVID-19. (A) Spearman correlations of cellular subset frequencies between indicated subsets, COVID score (1 = healthy, 2 = COVID-19 patient) and sex (1 = male, 2= female). (B) Top 5 correlations of subset frequencies with severity of infection scored as WHO ordinal scale (4 = moderate, 6= severe, 7 = critical). Edge width is proportional to the Spearman’s correlation strength. Edges connecting to other cellular subsets are positive correlations between indicated subsets. Layout by ForceAtlas2 using edge weights as input. Significance *P ≤ 0.05, **P ≤ 0.01, ***P ≤ 0.001 by Spearman’s rank correlation (A and B). Effector memory (EM), CM, terminal effector CD45RA positive (TEMRA). Expression values are arcsinh-transformed (cofactor: 5) dual counts.