Identification of a Pyroptosis-Related Gene Signature for Prediction of Overall Survival in Lung Adenocarcinoma

Abstract

Pyroptosis is a kind of programmed cell death that is characterized by inflammation. However, the expression of pyroptosis-related genes and their connection with prognosis in lung adenocarcinoma (LUAD) remain unknown. The aim of this study is to create and validate a LUAD prediction signature based on genes associated with pyroptosis. The TCGA and GEO were used to collect gene sequencing data and clinical information for LUAD samples. To identify patients with LUAD from the TCGA cohort, consensus clustering by pyroptosis-related genes was employed. Our prognostic model was constructed using LASSO-Cox analysis after Cox regression using differentially expressed genes. To predict patient survival, we created a seven-mRNA signature. Additionally, reliability and validity were established in the GEO cohort. To assess its diagnostic and prognostic usefulness, an integrated bioinformatics method was used. Using a risk score with varying overall survival (OS) in two cohorts (all $p<0.001$ ), a seven-gene signature was developed to categorize patients into two risk categories. The signature was shown to be an independent predictor of LUAD using multivariate regression analysis. The signature was linked to a variety of immune cell subtypes according to a study of immune cell infiltration. We constructed a signature consisting of seven genes as a robust biomarker with potential for clinical use in risk stratification and OS prediction in LUAD patients, as well as a potential indicator of immunotherapy in LUAD.