A novel epithelial-mesenchymal transition-related gene signature for prognosis prediction in patients with lung adenocarcinoma

Abstract

Traditional pathological diagnoses and clinical methods are insufficient to accurately predict the prognosis of lung adenocarcinoma (LUAD). Epithelial-mesenchymal transition (EMT) process is closely related to tumor cell migration. However, the prognostic value of EMT-related genes in LUAD is still unclear. In this study, we collected bulk RNA-sequencing (RNA-seq) and microarray data of LUAD patients from public databases and identified different expressed EMT-related genes in tumor and normal tissues. Then, we used the least absolute shrinkage and selection operator Cox regression model to develop a multigene signature in the cancer genome atlas (TCGA) cohort and validated the model in the OncoSG (Singapore Oncology Data Portal) cohort as well as other datasets. Finally, we constructed a 12-gene signature to divide LUAD patients into high-risk and low-risk groups of overall survival (OS), which has a better stability and accuracy in predicating the OS of patients compared with some other published signatures of LUAD. In addition, evaluation of the risk model using the time-related receiver operating characteristic (ROC) curve confirmed the predictive ability of the model. Functional analysis showed that these genes are related to immunity. CD8 T cell and CD4 T cell types were significantly negatively correlated with the risk score in the analysis of immune infiltration. In general, our model provides useful information that may help clinicians better predict the prognosis of LUAD patients and provides potential targets for immunotherapy of LUAD.

Keywords

• Epithelial-mesenchymal transition
• Lung adenocarcinoma
• Gene signature
• Risk model
• Immunity