Figure 2. Biological processes and pathway functional enrichment analysis of 371 high frequency/DE-mRNAs (HF/DE-mRNAs) extracted from 10 hepatocellular carcinomas (HCC) datasets. (A) The top 20 (sorted by P-value) from each enrichment analysis result for GO biological processes of HF/DE-mRNAs and the pathway enrichment analysis results for 3 data sources. (B) A clustering network formed by correlated terms from the functional enrichment analysis. (C) Protein-protein interaction (PPI) enrichment network of HF/DE-mRNA translation products constructed based on the molecular complex detection (MCODE) algorithm. (D) Gene set enrichment analysis (GSEA) of HF/DE-mRNAs in 10 datasets, showing the pathways with P < 0.05 and false detection rate (FDR) <0.25. The reference gene sets are the hallmark gene sets. In the TCGA-LIHC dataset, (E) the differences between the high hypoxia score group and the low hypoxia score group in 15 gene programs related to biological behaviors of tumors. These gene programs were identified by Hoadley's team, and the upregulation and downregulation trends were calculated as single-sample GSEA z-scores.