A new gene signature associated with disulfidptosis that forecasts myasthenia gravis and suggests infiltration of immune microenvironment in thymoma patients

Abstract

IntroductionThe mechanism of thymoma-associated myasthenia gravis (TAMG) is currently unknown, although patients with TAMG experience more severe myasthenic symptoms and have worse prognoses compared to regular thymoma patients. The objective of this research is to create a transcriptome map of TAMG using genes linked to disulfidptosis, detect possible biomarkers, and examine the disparities in the tumor immune microenvironment (TIME) among different thymoma patients. The findings will offer valuable knowledge for personalized treatment alternatives. MethodsThymoma samples' RNA-seq data, along with their corresponding clinical data, were acquired from the TCGA database using methods. Next, genes and disulfidptosis-related lncRNAs(DRLs) were chosen through correlation analysis. Then, a prediction model of TAMG was established by LASSO regression. Subsequent to that, an analysis of the mutation data, the tumor mutational burden (TMB), and the assessment of immune and stromal elements within the tumor microenvironment were conducted. ResultsA total of 87 patients diagnosed with thymoma were included in the study, with 29 of them having TAMG. We discovered a group of 325 lncRNAs in this sample that showed significant associations with genes related to disulfidptosis, with 25 of them displaying significantly altered expression. Moreover, utilizing LASSO regression, we constructed a predictive model incorporating 11 DRLs. The analysis revealed an area under the curve (AUC) of 0.934 (CI 0.879–0.989), a cut-off value of 0.797, along with a sensitivity of 82.8 % and specificity of 93.1 %. Furthermore, we examined the TIME in both the high-risk and low-risk groups, and observed noteworthy disparities in B cells, T cells, and APC among the two groups (p < 0.05). ConclusionThis research offers the initial examination of genes associated with disulfidptosis and TAMG through genomic and transcriptomic analysis. Furthermore, a strong risk forecasting model was created and the significance of TIME in TAMG was also clarified. The discoveries offer significant understanding into the molecular processes of TAMG and present possible indicators for categorizing risk.