Fig. 4

ComboSC uncovers the potential of tumor immune microenvironment recovery from immune cell exhaustion dynamics. a The heatmap shows trajectory scores of 44 middle immune score samples in four immune exhaustion cell trajectories. The gray color indicates that the trajectory does not exist in the sample. b Trajectory of active T cell exhaustion and memory T cell exhaustion. c Trajectory of tumor-associated macrophage (TAM) differentiation. d Trajectory of cancer-associated fibroblast (CAF) differentiation. e Drug-gene relationship used in comboSC, searched from the L1000CDS search engine. f The volcano plot identifies the differentially expressed genes along with the immune cell exhaustion trajectory in su008. Genes with p value < 10⁻⁵ and fold change > 2 are colored red. g Trajectory of T cell exhaustion in T15. h The volcano plot identifies the differentially expressed genes along with the immune cell exhaustion trajectory in T15. Genes with p value < 10⁻⁵ and fold change > 2 are colored red. i Drug effect of ivermectin to differentially expressed genes in Fig. 4h. The arrows and nocks between drug and genes represent the activation and suppression function of ivermectin to candidate genes, respectively. The red and blue colors of the gene labels represent upregulated and downregulated in the T cell exhaustion trajectory of T15, respectively. j This bubble chart shows the count and enrichment significance of the top 10% comboSC predictions registered in the ClinicalTrials.gov database. The x-axis is the fold enrichment of clinically validated drugs in the top 10% comboSC predictions compared to the ratio (60/1280) of validated drugs in all CMap drugs (Additional file 2: Table S4), and the y-axis is the dataset name and sample ID (Additional file 2: Table S2). The color represents the p value from the enrichment analysis, and the size represents the number of clinically validated drugs in the top 10% comboSC predictions