Abstract
Liver cancer research highlights the kinome's critical role in disease initiation and progression. However, comprehensive data analysis on the kinome's impact on hepatocellular carcinoma (HCC) prognosis is limited. We used the TCGA-LIHC mRNA expression profiles, analyzing them with various R packages. Key methods included univariate Cox regression for prognostic gene identification, consensus clustering for subtype classification, Gene Set Enrichment Analysis (GSEA), and immune landscape evaluation. A prognostic model was developed using LASSO Cox regression, and chemotherapy drug sensitivity was assessed using the pRRophetic package. We identified 45 kinases-related differentially expressed genes (DEGs), with 27 linked to HCC prognosis. Cluster analysis divided these genes into two subtypes, with distinct prognoses. We discovered 157 DEGs between kinase-related subtypes, 120 of which were prognostically relevant. A kinase-related gene signature (KRS) was developed for prognostic prediction. The high-KRS group showed poorer survival in TCGA-LIHC and validation cohorts, with notable differences in immune cell infiltration and checkpoint gene expression. This group also showed varying sensitivity to common drugs and anti-PD-L1 treatment. In contrast, the low-KRS group might respond better to anti-PD-1 immunotherapy. Our study introduces a kinase-related gene signature as a novel tool for predicting HCC prognosis. This signature aids in tailoring personalized treatment strategies, potentially improving clinical outcomes in HCC patients.