Bioinformatics analysis of a-three-gene signature as an independent prediction of survival in follicular gastritis developing into gastric cancer

From follicular gastritis (FG) to gastric cancer (GC), the expression of genes has changed and the function of these genes has not been elucidated. Data of FG samples and GC samples were downloaded from GSE106656 and GSE116312 in GEO database. Clinical data of GC were downloaded from TCGA database. The common differential expression genes (DEGs) were enriched and analyzed by GO and KEGG in David. The interaction information of proteins was obtained from the STRING database and the core genes in the Cytoscape software were excavated. After KEGG, PPI and KM survival analysis of DEGs, the significant genes were intersected to fulfill the predicted genes. Univariate/Multivariate Cox regression analysis were used to get the prediction signatures. 96 up-regulated genes were obtained from the data set. Three genes, CAV1 COL12A1 and ANGPT2, were acquired by survival analysis screening. Age, gender, lymph node status, and the 3-gene signature were selected as independent survival variables in the nomogram. The nomogram showed that C-index was 0.68 (95% CI, 0.60 to 0.75). Our study identified the proposed nomogram may be used to predict the prognosis of patients from FG to GC, which may provide a potential useful guidance for preventive and epidemiological applications in GC. •The change of gene expression from follicular gastritis to gastric cancer was described.•CAV1, COL12A1 and ANGPT2 were proposed as new prognostic factors.•The 3-gene signature was selected as independent survival variables in the nomogram.