The study analyzed 30 samples of peripheral blood mononuclear cells (15 from patients with hemophagocytic lymphohistiocytosis - HLH - and 15 controls) using mRNA sequencing to identify genes related to iron metabolism. From 464 differentially expressed genes and 520 iron metabolism genes, 10 candidates emerged, from which machine learning selected five key genes: ALOX15, CAT, HBZ, MT2A and CYGB. These genes are related to antioxidant activity, response to toxic substances and detoxification of inorganic compounds, including oxidative phosphorylation pathways and Huntington's disease. ALOX15 and CAT significantly correlate with activated dendritic cells and memory CD4 T cells; ALOX15 has a positive correlation with HBZ (r > 0.3, p < 0.05) and a negative correlation with CYGB and MT2A. Drug prediction showed 15 compounds targeting ALOX15, MT2A and CAT. RT-qPCR confirmed down-regulation of ALOX15, CAT and HBZ and up-regulation of MT2A and CYGB in HLH patients.