Fig. 4: The role of transcriptional regulators in progression of MASLD.

a Correlation between all identified transcriptional regulators and patient characteristics, ordered by the sum of correlation to Kleiner fibrosis grade, NAFLD activity score, and SAF diagnosis. Cutoff was arbitrarily selected. b Hierarchical clustering of above cutoff transcriptional regulators with their correlation to the 234 significantly differentially expressed proteins of Kleiner fibrosis grade and MASLD. The columns (234 proteins) were first hierarchically clustered (Supplementary Fig. S9a), generating three clusters and then subsequently sorted within each of these three clusters by their cell type cluster (Fig. 2a). Proteins found in both the row and column (perfect correlation) has been grayed out. Squares indicate proteins with a higher correlation and were primarily chosen based on their correlation to the highest-ranking transcriptional regulators. The full figure containing all protein names can be found in Supplementary Fig. S9b. c–e Pathway enrichment analysis based on the Reactome pathway database for proteins in each of the column-wise clusters. f, g Fold change in protein expression for a selected subset of transcriptional regulators according to increasing f SAF diagnosis and g Kleiner fibrosis grade. Boxplot are shown in the style of Tukey (median, hinges: Q1 and Q3, and whiskers: 1.5x of IQRs from Q1 and Q3) and statistical significance calculated using Wilcoxon signed-rank test and BH correction method. SAF steatosis, activity, and fibrosis, NAFLD non-alcoholic fatty liver disease, IMX immune cell mixture, MMC macrophages and monocytes, HEP hepatocytes, HSC hepatic stellate cell, LEC liver endothelial cells, BH Benjamini–Hochberg.