Integrative multi-omics and network toxicology reveal candidate molecular targets of deoxynivalenol in melanoma
摘要
Deoxynivalenol (DON) has been reported to exhibit skin toxicity and carcinogenic potential; however, its effects on melanoma remain unclear. We integrated network toxicology, single-cell transcriptomic analysis, molecular docking, and publicly available Human Protein Atlas immunohistochemical evidence to investigate the potential toxicological effects of DON on melanoma and identify candidate molecular targets. A total of 5283 DON-related genes were identified from the SwissTargetPrediction, CTD, and SEA databases, and 5101 melanoma-related genes were obtained from the GeneCards, TTD, and OMIM databases. Differentially expressed genes from the GSE15605 dataset were used for intersection validation. After further screening with three machine learning algorithms, MYEF2 was identified as a core target potentially involved in DON-induced melanoma. ROC analysis showed that MYEF2 achieved AUC values above 0.75 in both the training and validation cohorts. Single-cell transcriptomic analysis revealed the expression and distribution of MYEF2 across melanoma cell subtypes. Immune infiltration analysis further showed significant associations between MYEF2 and multiple immune cell populations. Molecular docking supported a stable interaction between DON and MYEF2, and HPA validation confirmed its elevated expression in melanoma. These findings suggest that MYEF2 may represent a candidate molecular target involved in DON-associated melanoma progression, providing a theoretical basis for future studies on the mechanisms and risk assessment of DON exposure.