In silico prediction of paradoxical effect for oxaliplatin in gastric cancer patients based on their transcriptomic profile.

Gastric cancer (GC) is a major global health concern, ranking as the fifth most commonly diagnosed cancer. New treatment strategies like chemoprevention with oxaliplatin (OXA) are emerging, but safety data for GC patients are limited. This in silico study aimed to predict potential paradoxical effects of OXA treatment in GC patients using computational analysis. RNA-sequencing data from GSE26942, GSE66229, and TCGA-STAD datasets were analyzed. Differential gene expression was identified using GEO2R and DESeq2. Pathway enrichment and protein-protein interaction networks were constructed to pinpoint genes crucial for GC progression. Finally, the R Survival package identified survival-related differentially expressed genes (DEGs). Interactions between OXA and GC-related genes were retrieved from the CTD database and compared with DEGs. A total of 151 dysregulated genes were identified across the datasets, comprising 112 downregulated and 39 upregulated genes. Thirteen genes emerged as potential prognostic biomarkers for overall survival. OXA interacted with 97 genes, of which 14 were linked to both OXA and differentially expressed genes in GC. OXA potentially reversed the expression of seven genes associated with GC progression (BIRC5, CAV1, CDH2, IL6, JUN, SERPINB2, TYMS), while promoting the expression of six others (BLVRB, CDKN2A, MAPK3, PLAU, PTGS2, SERPINE1). Notably, SERPINE1 showed a strong correlation with overall survival. Our findings suggest that a patient's genetic profile, particularly SERPINE1 expression levels, might be crucial for determining the safety and efficacy of OXA treatment for GC.