Extended Data Fig. 5: Cross-participant module discovery recapitulates aspects determined by calculating participant-specific modules.
From: Integrated single-cell analysis of multicellular immune dynamics during hyperacute HIV-1 infection
a, Timepoints chosen for testing significant changes in module expression over time across all participants. Timepoints were chosen based on peak expression of modules in MM1 and MM3. We used an ANOVA model to account for participant-specific features. b, Significant cross-participant modules that map to participant-specific modules (share at least 5 genes), separated by cell type. Median module expression is plotted for each module split by participant. Error bars depict the upper and lower quartiles for all cells across all four individuals at each time point. Boxed modules demonstrate consistent directional trends in expression between each pair of timepoints in at least 3/4 participants. P-I = pre-infection; P = peak; P-P = post-peak; 1-Y = 1 year. See a and Supplementary Table 2 for cell numbers per cell type per participant. c, Significant cross-participant modules that do not correspond to participant-specific modules. d, Sankey diagram demonstrating the gene overlap between participant specific modules (left) and cross-participant modules (right). Node size correlates with the number of genes within the module and edge width correlates with the number of shared genes between modules. Only overlaps consisting of ≥5 genes have edges depicted.
