An increasing number of studies have demonstrated the pivotal role of microbiota changes in the onset, progression, diagnosis, treatment, and prognosis of lung adenocarcinoma (LUAD). However, a comprehensive analysis of intratumoral microbiome variation across distinct LUAD stages has not been performed. The aim of this study was to identify the microbial markers that significantly vary during tumor stage of LUAD.
Here, we used the cancer genome atlas (TCGA) database to comprehensively compare and analyze the differences in microbial composition between 267 patients with early and 224 patients with advanced LUAD. In order to determine the best biomarkers, we used the random forest (RF) model and found that the microbial markers have a certain ability in predicting the stage of LUAD.
We found that there were certain differences in the microbiome of patients with LUAD at different stages, especially in the tumor tissues of patients with advanced LUAD, whose co-abundance network was significantly more complex. We also found that five bacterial biomarkers (
Our study revealed the microbial profile of patients with LUAD and the intrinsic pathogenic mechanism between the microbiome and the disease, and established a multi-omics model to determine LUAD tumor stage.