Inconsistency in the expression pattern of a five-lncRNA signature as a potential diagnostic biomarker for gastric cancer patients in bioinformatics and in vitro

Due to diagnosis of gastric cancer in advanced stages as well as its poor prognosis, finding biomarkers is essential. In this study, using the TCGA RNAseq data of gastric cancer patients, we evaluated the diagnostic value of lncRNAs that had differential expression. We evaluated P-value, FDR, and log fold change for whole transcripts. Next, by comparison of the RNAseq gene names with the total known lncRNA names, we identified differential expressed lncRNAs. Following this, specificity and sensitivity for lncRNAs coming from the previous step were calculated. For more confirmation, we predicted target genes and performed GO and KEGG signaling pathway analysis.  In the end, we examined reliability and consistency of expression of this signature in three gastric cancer cell lines and one of them in twenty tumors and tumor-adjacent normal tissue samples using qRT-PCR. Five lncRNAs had proper sensitivity and specificity and had target genes involved in cancer-related signaling pathways; however, they showed different expression patterns in TCGA data and in vitro. The results of our study demonstrated that the five-lncRNAs PART1, UCA1, DIRC3, HOTAIR, and HOXA11AS require more investigation to be confirmed as diagnostic biomarkers in gastric cancer.