Bacterial lipopolysaccharide related genes signature as potential biomarker for prognosis and immune treatment in gastric cancer

Abstract The composition of microbial microenvironment is an important factor affecting the development of tumor diseases. However, due to the limitations of current technological levels, we are still unable to fully study and elucidate the depth and breadth of the impact of microorganisms on tumors, especially whether microorganisms have an impact on cancer. Therefore, the purpose of this study is to conduct in-depth research on the role and mechanism of prostate microbiome in gastric cancer (GC) based on the related genes of bacterial lipopolysaccharide (LPS) by using bioinformatics methods. Through comparison in the Toxin Genomics Database (CTD), we can find and screen out the bacterial LPS related genes. In the study, Venn plots and lasso analysis were used to obtain differentially expressed LPS related hub genes (LRHG). Afterwards, in order to establish a prognostic risk score model and column chart in LRHG features, we used univariate and multivariate Cox regression analysis for modeling and composition. In addition, we also conducted in-depth research on the clinical role of immunotherapy with TMB, MSI, KRAS mutants, and TIDE scores. We screened 9 LRHGs in the database. We constructed a prognostic risk score and column chart based on LRHG, indicating that low risk scores have a protective effect on patients. We particularly found that low risk scores are beneficial for immunotherapy through TIDE score evaluation. Based on LPS related hub genes, we established a LRHG signature, which can help predict immunotherapy and prognosis for GC patients. Bacterial lipopolysaccharide related genes can also be biomarkers to predict progression free survival in GC patients.