Skip to main content
. 2026 Jul 13;17:1843796. doi: 10.3389/fimmu.2026.1843796

Figure 3.

Four panels labeled a, b, c, and d, each displaying a heatmap of gene expression data with samples grouped by color-coded experimental conditions, alongside a principal component analysis (PCA) scatter plot summarizing group clustering for HepG2, IFN_LPS, and IL10 macrophages according to phenotypic and temporal groupings.

Comparison of markers selected by established methods and the CDBGE selector for their application in macrophage profiling. Top panel: Heatmaps of markers selected to characterize different types of macrophages using (A) CDBGE, (B) conventional, (C) Random Forest and (D) WGCNA methods. Performance is illustrated by comparing the true group of each sample (Real group, second row in heatmap) with the phenotype assigned based on the selected markers (third row in heatmap). Lower panel: PCA plots of macrophage profiling using the methods indicated above. Dim, dimension.