Development and Validation of a Novel Circadian Rhythm-Related Signature to Predict the Prognosis of the Patients with Hepatocellular Carcinoma

Figure 1

The flow diagram of the integral analysis. Gene expression profiles and corresponding clinical data were obtained from TCGA and GTEx for tumor and normal tissues, excluding samples with incomplete information. Differential expression analysis was performed to identify 53 differentially expressed circadian rhythm-associated genes. Next, HCC patients in TCGA dataset were classified into two distinct molecular subtypes (C1 and C2 clusters) by consistent clustering analysis of these 53 genes. The two HCC molecular subtypes differed significantly in terms of clinicopathological characteristics and overall survival. We then used Cox regression analysis and LASSO regression to build a risk model based on six circadian rhythm-related genes, and we used TCGA cohort and the GEO cohort as the training and validation sets for the model, respectively. Finally, variance analysis, enrichment analysis, and ssGSEA were performed based on the risk model.