Identification of ferroptosis-related genes associated with diffuse large B-cell lymphoma via bioinformatics and machine learning approaches

Ferroptosis has emerged as a critical mechanism in the development and progression of various tumors, particularly diffuse large B-cell lymphoma (DLBCL). However, the thorough characterization of ferroptosis-related genes in DLBCL remains inadequately explored. We retrieved datasets associated with DLBCL and ferroptosis gene sets from the Gene Expression Omnibus (GEO) database and the Ferroptosis Database (FerrDb), resulting in the identification of 27 differentially expressed ferroptosis-related genes (DE-FRGs) linked to DLBCL. Utilizing the LASSO and Support Vector Machine Recursive Feature Elimination (SVW-RFE) algorithms, we identified 10 genes-MT1G, MTOR, BRD4, ACO1, SAT1, PEBP1, LPIN1, ATM, SRXN1, and PRDX1-as key biomarker candidates with significant diagnostic potential. Functional enrichment analyses revealed that these biomarker genes are likely involved in regulating several critical biological pathways implicated in DLBCL pathogenesis, including immune response, oxidative phosphorylation, and cell cycle regulation. Moreover, we identified 246 potential therapeutic agents targeting these 10 biomarker genes. Concurrently, competitive endogenous RNA (ceRNA) network analysis uncovered a complex regulatory network centered on the identified biomarker genes. Additionally, CIBERSORT analysis highlighted notable alterations in the immune microenvironment of DLBCL patients. We propose a diagnostic strategy that provides novel insights into the molecular mechanisms underlying DLBCL. Nevertheless, further validation of the practical value of this strategy for DLBCL diagnosis is necessary before its clinical application.