Identification and characterization of a novel molecular classification incorporating oxidative stress and metabolism-related genes for stomach adenocarcinoma in the framework of predictive, preventive, and personalized medicine.

Background: Stomach adenocarcinoma (STAD) is one of the primary contributors to deaths that are due to cancer globally. At the moment, STAD does not have any universally acknowledged biological markers, and its predictive, preventive, and personalized medicine (PPPM) remains sufficient. Oxidative stress can promote cancer by increasing mutagenicity, genomic instability, cell survival, proliferation, and stress resistance pathways. As a direct and indirect result of oncogenic mutations, cancer depends on cellular metabolic reprogramming. However, their roles in STAD remain unclear. Method: 743 STAD samples from GEO and TCGA platforms were selected. Oxidative stress and metabolism-related genes (OMRGs) were acquired from the GeneCard Database. A pan-cancer analysis of 22 OMRGs was first performed. We categorized STAD samples by OMRG mRNA levels. Additionally, we explored the link between oxidative metabolism scores and prognosis, immune checkpoints, immune cell infiltration, and sensitivity to targeted drugs. A series of bioinformatics technologies were employed to further construct the OMRG-based prognostic model and clinical-associated nomogram. Results: We identified 22 OMRGs that could evaluate the prognoses of patients with STAD. Pan-cancer analysis concluded and highlighted the crucial part of OMRGs in the appearance and development of STAD. Subsequently, 743 STAD samples were categorized into three clusters with the enrichment scores being C2 (upregulated) > C3 (normal) > C1 (downregulated). Patients in C2 had the lowest OS rate, while C1 had the opposite. Oxidative metabolic score significantly correlates with immune cells and immune checkpoints. Drug sensitivity results reveal that a more tailored treatment can be designed based on OMRG. The OMRG-based molecular signature and clinical nomogram have good accuracy for predicting the adverse events of patients with STAD. Both transcriptional and translational levels of ANXA5, APOD, and SLC25A15 exhibited significantly higher in STAD samples. Conclusion: The OMRG clusters and risk model accurately predicted prognosis and personalized medicine. Based on this model, high-risk patients might be identified in the early stage so that they can receive specialized care and preventative measures, and choose targeted drug beneficiaries to deliver individualized medical services. Our results showed oxidative metabolism in STAD and led to a new route for improving PPPM for STAD.