Lung cancer ranks among the most prevalent and lethal cancers globally. Specifically, lung adenocarcinoma (LUAD) is
a dominant subtype. While there have been substantial advancements in LUAD diagnosis and treatment, our current
prognostic evaluation methods are still suboptimal. Although the disulfidptosis (DS) pathway has been implicated in
several cancers, its association with genes affecting LUAD prognosis remains ambiguous. Our study aspires to formulate
a LUAD prognosis model rooted in genes tied to the dual sulfur pathway, shedding light on its underlying biological
mechanisms.
To achieve this, we sourced sequencing data related to lung adenocarcinoma from public repositories like The Cancer
Genome Atlas (TCGA) and Gene Expression Omnibus (GEO)database. Using the gsva method, we filtered for genes
prominently active within the disulfidptosis pathway from the TCGA dataset. We used lossa regression and random forest
methods to identify feature genes, and established a feature genes model using multivariate regression.Using the feature
genes, we determined the risk values of each sample and categorized them into high-risk and low-risk groups. Following
this, we carried out differential analysis based on the high and low-risk groups, and using the results from the differential
analysis, we employed the Gene set variation analysis (GSVA) method to analyze gene expression data to identify which
the kyoto encyclopedia of genes and genomes (KEGG) pathway and gene ontology (GO) analysis of biologic process
(BP) pathways were significantly enriched. At the same time, we explored the tumor immune microenvironment, gene
mutations in tumors, and drug susceptibility. We also devised and corroborated a gene model through an external dataset.
Lastly, we predicted the 1-, 3-, and 5-year survival rates using feature genes and clinical information.
Our analytical efforts pinpointed seven feature genes: SLC2A1, LDHA, SNAI2, ACO2, FGF12, ANP32B, and ST13,
which were recognized as risk determinants. A gene model was framed and its predictive efficacy was tested using an
external data set. A marked disparity was found between the survival rates of high and low-risk groups (P¡0.05). Both
univariate and multivariate Cox regression analysis corroborated that genes linked to the dual sulfur pathway stand
as vital independent prognostic indicators, unaffected by clinical attributes. Using GSEA, we undertook enrichment
examinations of KEGG and GOBP pathways for the high and low-risk clusters. This highlighted that genes with varied
expressions were chiefly operational in pathways such as P53-SIGNALING-PATHWAY, CELL-CYCLE, SPLICEOSOME,
and DNA-REPLICATION. The high-risk cluster showed enhanced associations with Myeloid dendritic cells, T cell
CD4+ (non-regulatory), Macrophage M1, and uncharacterized cells when analyzing immune cell infiltration. Moreover,
a noticeably elevated gene mutation frequency was observed in the high-risk group. By integrating risk scores from genes
linked to the dual sulfur pathway and three clinical parameters, we proposed a predictive model to estimate 1-year, 3-year,
and 5-year survival probabilities for LUAD patients.
Conclusively, leveraging feature genes from the disulfidptosis pathway offers a promising avenue for forecasting lung
adenocarcinoma patient outcomes and gauging their tumor immune microenvironment. This could equip medical
practitioners with insightful tools for diagnosing and addressing LUAD.
