Discovery and validation of ferroptosis-related molecular modules and immune signatures in epilepsy

Abstract

The pathophysiology of epilepsy is still not fully understood. Though little is known about the molecular immunological mechanisms underlying ferroptosis, numerous lines of evidence point to its critical role in the pathophysiology of epilepsy. Thus, the objective of this work was to thoroughly examine and analyze the molecular mechanism and immunological features of genes connected to ferroptosis in the pathophysiology of epilepsy. For our research, we downloaded ferroptosis-related gene sets from FerrDb and got blood and brain tissue datasets for epilepsy from the GEO database. The most pertinent Hub gene for epilepsy was found using two machine learning algorithms: Random Forest (RF) and Multiclass Support Vector Machine Recursive Feature Elimination (mSVM-RFE). There are two sections to the Hub gene research. Part I: Immunological features of various clusters were analyzed and epilepsy patients' genotypes were determined using unsupervised cluster analysis. The PCA method was used to quantify the FRGscore by analyzing the connection between FRGscore and patient clinical information. Part II: By combining methods (GSEA, GSVA, and CIBERSORT), we clarify the biological processes associated with Hub genes and their roles in the immune milieu. Logistic regression models were utilized for additional analysis of hub genes. Lastly, RT-qPCR was used to confirm the Hub gene's expression in the brain tissue of mice given KA to induce epilepsy. We were able to identify three Hub genes in total using two machine learning algorithms. Immune infiltration study revealed that the level of immune infiltration in type B was much higher than that in type A, suggesting that type B may be at the pinnacle of neuroinflammation in epilepsy. Unsupervised clustering successfully identified two separate clusters. Second, the nomo diagram and logistic regression technique were used to build the epilepsy diagnosis model. Ultimately, the Hub gene results from the RT-qPCR validation were in agreement with the findings of the bioinformatic analysis, demonstrating the accuracy of the data. Our research offers fresh perspectives on the roles played by immunological systems and ferroptosis-related molecular patterns in epilepsy. It also establishes a theoretical framework for the potential addition of additional epilepsy diagnostic markers.