Dissecting Immune Cell Stat Regulation Network Reveals Biomarkers to Predict ICB Therapy Responders in Melanoma

Immunotherapy is a revolutionary strategy in cancer therapy, but the resistance of which is one of the important challenges. Detecting the regulation of immune cells and biomarkers concerning immune checkpoint blockade (ICB) therapy is of great significance. Here, we firstly constructed regulation networks for 11 immune cell clusters by integrating biological pathway data and single cell sequencing data in metastatic melanoma with or without ICB therapy. We then dissected these regulation networks and identified differently expressed genes between responders and non-responders. Finally, we trained and validated a logistic regression model based on ligands and receptors in the regulation network to predict ICB therapy response. We discovered the regulation of genes across eleven immune cell stats. Functional analysis indicated that these stat-specific networks consensually enriched in immune response corrected pathways and highlighted antigen processing and presentation as a core pathway in immune cell regulation. Furthermore, some famous ligands like SIRPA, ITGAM, CD247and receptors like CD14, IL2 and HLA-G were differently expressed between cells of responders and non-responders. A predictive model of gene sets containing ligands and receptors performed accuracy prediction with AUCs above 0.7 in a validation dataset suggesting that they may be server as biomarkers for predicting immunotherapy response. In summary, our study presented the gene–gene regulation landscape across 11 immune cell clusters and analysis of these networks revealed several important aspects and immunotherapy response biomarkers, which may provide novel insights into immune related mechanisms and immunotherapy response prediction.