Analyses of single-cell and bulk RNA-sequencing datasets identify biomarker-based molecular subtypes and a prognostic signature in lung adenocarcinoma.

Background: Immune checkpoint blockade has improved outcomes for a subset of patients with lung adenocarcinoma (LUAD), yet primary and acquired resistance remain common. Given the heterogeneity of the tumor microenvironment (TME), biomarkers that reflect both tumor-intrinsic programs and immune context are needed for prognostic stratification and for generating clinically relevant hypotheses. The aim of this study was to identify immune-associated genes and construct a prognostic signature for LUAD by integrating single-cell and bulk transcriptomic data. Methods: We integrated single-cell RNA sequencing (scRNA-seq; GSE131907) with bulk transcriptomic profiles from The Cancer Genome Atlas (TCGA)-LUAD to identify immune-associated genes and construct a prognostic signature. Major cell populations were delineated and annotated using reference-based methods, and immune-associated differentially expressed genes (DEGs) were derived from the scRNA-seq analysis. Genes consistently dysregulated in both scRNA-seq and bulk data were carried forward for functional enrichment, survival analyses, and molecular subtype discovery. A multigene risk score was built using Cox regression and validated in an independent Gene Expression Omnibus (GEO) cohort (GSE13213) and evaluated using Kaplan-Meier analysis, time-dependent discrimination metrics, and multivariable adjustment for clinical covariates. Immune infiltration and immune functional states were characterized using Estimation of STromal and Immune cells in MAlignant Tumor tissues (ESTIMATE), single-sample gene set enrichment analysis (ssGSEA), Cell-type Identification By Estimating Relative Subsets Of RNA Transcripts (CIBERSORT), and Tumor Immune Dysfunction and Exclusion (TIDE). Protein-level evidence was examined using the Human Protein Atlas (HPA), and messenger RNA (mRNA) expression was assessed in paired clinical specimens by reverse transcription quantitative polymerase chain reaction (RT-qPCR) (n=10 pairs). Results: Across datasets, we identified immune-associated genes consistently altered in LUAD and enriched for antigen presentation and immune response pathways. A four-gene signature ( Conclusions: Our integrative analysis nominates a four-gene immune-associated signature that captures prognostic heterogeneity in LUAD and is linked to distinct TME states. These findings provide candidate biomarkers and testable hypotheses for immunotherapy-related mechanisms, which require validation in independent immunotherapy-treated cohorts and functional experiments.