Identifying Potential Diagnostic and Therapeutic Targets for Infantile Hemangioma Using WGCNA and Machine Learning Algorithms

Infantile hemangioma (IH) is the most common benign vascular tumor during infancy and childhood and is characterized by abnormal vascular development. It is the most common vascular tumor and its related mechanisms and treatments remain a problem. IH-related biomarkers have been identified using transcriptome analysis and can be used to predict clinical outcomes. This study aimed to identify the key target genes for IH treatment and explore their possible roles in the IH pathophysiology. Gene records were acquired from the Gene Expression Omnibus database. Utilizing integrated weighted gene co-expression network examination, gene clusters were determined. Single-sample gene set enrichment analysis was performed to gauge immune infiltration. Essential genes were identified via Random Forest and Least Absolute Selection and Shrinkage Operator analyses. Ultimately, a set of five pivotal genes associated with the ailment was identified (NETO2, IDO1, KDR, MEG3, and TMSB15A). A nomogram for predicting IH diagnosis was constructed based on hub genes. The calibration curve showed valid agreement between the prediction and conclusion that the key genes in the model were clinically significant. Neuropilin and Tolloid-like 2 (NETO2) are closely associated with tumor development. The role value of NETO2 expression levels increased in hemangioma-derived endothelial cells (HemECs). After silencing NETO2, the growth and migration of cancer cells were significantly restrained. This study revealed the critical role of NETO2 in IH development, suggesting that targeting NETO2 may be effective in improving the therapeutic outcome of IH.