Metabolism is an ordered series of biological processes that occur in an organism. Altered cellular metabolism is often closely associated with the development of cancer. The aim of this research was to construct a model by multiple metabolism-related molecules to diagnose and assess the prognosis of patients.
WGCNA analysis was used to screen out differential genes. GO, KEGG are used to explore potential pathways and mechanisms. The lasso regression model was used to filter out the best indicators to construct the model. Single-sample GSEA (ssGSEA) assess immune cells abundance, immune terms in different Metabolism Index (MBI) groups. Human tissues and cells were used to verify the expression of key genes.
WGCNA clustering grouped genes into 5 modules, of which 90 genes from the MEbrown module were selected for subsequent analysis. GO analysis was found that BP mainly has mitotic nuclear division, while KEGG pathway is enriched to Cell cycle, Cellular senescence. Mutation analysis revealed that the frequency of TP53 mutations was much higher in samples from the high MBI group than in the low MBI group. Immunoassay revealed that patients with higher MBI have higher macrophage and Regulatory T cells (Treg) abundance, while NK cells were lowly expressed in the high MBI group. RT-qPCR and immunohistochemistry (IHC) revealed that the hub genes expression is higher in cancer tissues. The expression in hepatocellular carcinoma cells was also much higher than that in normal hepatocytes.
In conclusion, a metabolism-related model was constructed that can be used to estimate the prognosis of hepatocellular carcinoma, and the clinical treatment of different hepatocellular carcinoma patients with medications was guided.