feature selection algorithms combined with experimental validation reveal col10a1 and tmprss4 genes as potential diagnostic biomarkers in pancreatic ductal adenocarcinoma

Background: pancreatic adenocarcinoma (paad) is often diagnosed at a late stage, preventing curative surgery. early detection is crucial for improving patient outcomes. this study aims to discover potential biomarkers for identifying asymptomatic paad tumors.method: in this case-control study, two gene expression datasets of paad and normal samples were collected from geo and tcga databases. independent analyses of these datasets were conducted, leading to the identification of genes common to both datasets. gene ontology and pathway enrichment analyses for the feature genes were conducted. following our strict criteria, three feature genes for experimental validation were selected. the reliability of the selected feature genes was determined through quantitative real-time polymerase chain reaction (qrt-pcr). data were analyzed using graphpad prism 8 software, employing the mann-whitney test and unpaired t-test.  a p-value of <0.05 was considered statistically significant.results: a total number of 33 genes common to both geo and tcga datasets were identified. gene ontology and pathway enrichment methods revealed that the selected genes were primarily associated with proteolysis and extracellular matrix organization. based on our criteria, three feature genes (col10a1, cthrc1, and tmprss4) were selected for experimental validation. the results of qrt-pcr on independent patient samples demonstrated that the expression levels of col10a1 and tmprss4 were significantly upregulated in paad tissues as compared with normal pancreatic tissues. in contrast, cthrc1 expression levels did not change significantly in paad in comparison with normal samples.conclusion: our findings suggest that col10a1 and tmprss4 can be attractive biomarkers for the mrna-based diagnosis of paad.