LDLRAD4 is a potential diagnostic and prognostic biomarker correlated with immune infiltration in myelodysplastic syndromes

Purpose Myelodysplastic syndromes (MDS) are a group of hematological disorders that remain relatively under-explored, which are characterized by inconspicuous early symptoms and generally poor prognosis. Owing to the complex and variable pathogenesis of MDS, there is a relative paucity of available therapeutic options. Consequently, in-depth investigation into the pathogenesis of MDS and the search for effective targeted therapies have become urgent priorities. Methods In this study, we leveraged the Gene Expression Omnibus (GEO) database to identify differentially expressed genes (DEGs) and conducted functional enrichment analysis. Utilizing three machine learning algorithms—Least Absolute Shrinkage and Selection Operator (LASSO), Support Vector Machine Recursive Feature Elimination (SVM-RFE), and Random Forest (RF)—we pinpointed hub genes. Furthermore, this study explored the relationship between hub gene expression levels and immune infiltration. Results Our analysis identified three hub genes: LDLRAD4, FAM43A, and KCNK5, with LDLRAD4 showing a close association with TGF-β and MAPK signaling pathways. Furthermore, this study revealed a positive correlation between LDLRAD4 expression levels and immune infiltration, particularly with natural killer (NK) cells, offering a novel immunological perspective on LDLRAD4. Ultimately, we observed that the overexpression of LDLRAD4 can suppress the proliferative capacity of MDS cells, induce cell cycle arrest, and enhance apoptosis. Conclusion We conclude that LDLRAD4, FAM43A, and KCNK5 are potential biomarkers for MDS. LDLRAD4’s overexpression in vitro inhibits MDS cell proliferation and promotes apoptosis, suggesting significant potential for immunotherapy research. These findings collectively identify LDLRAD4 as a promising therapeutic target for MDS. However, its clinical applicability warrants further investigation to validate its potential.