Background
Gastric cancer (GC)remains the third leading cause of cancer-related death. Cuproptosis has a high correlation with cancer development and progression, while Cuproptosis-related genes (CRGs) are rarely reported in GC. The aim of this multi-omics study was to investigate the prognostic value and biological functions of CRGs in GC, which may help guide precision medicine-based decision-making in GC patients.
Methods
RNA sequencing (RNA-seq) data, Copy number variations (CNV) data, and Single nucleotide variant (SNV) data were obtained from The cancer genome atlas (TCGA) database. Chi-squared test was adopted to screen differentially expressed CRGs (DE-CRGs) between samples from 14 kinds of carcinoma and adjacent tissue samples. Then, gastric cancer (GC) samples were divided into high- and low-expressed groups based on DE-CRGs for further overall survival (OS), progression-free survival (PFS), disease-free survival (DFS), and disease-special survival (DSS) analysis. After single-gene Receiver operating characteristic (ROC) analysis, biomarkers of GC was obtained eventually. Besides, methylation sites related with biomarkers were acquired and survival analysis was performed based on those sites. Next, the correlation between immune cells and biomarkers was verified. Finally, we established miRNA-mRNA, TFs-mRNA, and co-expression networks to detect factors that have a regulating effect on biomarkers.
Results
Four DE-CRGs including CDKN2A, DLD, GLS, LIAS, and PDHB in most of the 14 cancers were screened out. Seven CRGs including GLS, LIAS, CDKN2A, DLD, LDAT, MTF1 and PDHA1 have a significant difference in the survival of GC patients. Next, single-gene ROC proved that PDHB, CDKN2A, LIAS, and GLS significantly correlate with GC prognosis. Three CRGs including LIAS, GLS, and CDKN2A were remain as biomarkers based on the results we got previously, and were used to generate a nomogram. After, 3 methylation sites with a significant survival relationship which include cg13601799, 07562918, and 07253264 were found. Then, we found that B cells native is significantly correlated with CDKN2A, 4 immune cells such as T cells regulatory (Tregs) are significantly correlated with GLS, and 2 immune cells such as T cells CD4 memory activated are significantly correlated with LIAS. Moreover, we found 10 miRNA in the miRNA-mRNA network and 3 TFs in the TFs-mRNA network have a significant correlation with OS. Finally, 20 enrichment functions were obtained such as cardiac septum development, collagen fibril organization, and sensory organ morphogenesis on the basis of the co-expression network.
Conclusions
3 biomarkers with a prognosis prediction value of GC were found, and multi-factor regulatory networks was constructed to screen out 13 factors with regulating influences of biomarkers.