Development and Validation of a Novel Diagnostic Model for Childhood Autism Spectrum Disorder Based on Ferroptosis-Related Genes

Abstract

Autism spectrum disorder (ASD) is a highly genetic heterogeneous neurodevelopmental disorder, which is usually considered a heritable and heterogeneous neurodevelopmental disorder and has caused a great burden to society and families. Emerging roles of ferroptosis have been observed in neurological disorders. This study aimed to construct a diagnostic model based on ferroptosis-related genes (FRGs) to contribute to the early and precise diagnosis of childhood ASD. In the candidate FRGs, we identified 27 differentially expressed genes (DEGs) between ASD patients and typically developing (TD) controls. Four key FRGs were identified using the random forest analysis for further analysis. Utilization of the four gene expression, we constructed a diagnostic model and the AUC value in the training dataset (GSE18123) is 0.7002. We deem that a patient with a score less than 0.9904 is likely to have ASD. Three validation datasets (GSE111176, GSE113834, and GSE28521) were collected and the AUC value is 0.7442, 0.7444, and 0.6474, respectively. A multi-factor regulatory network based on four FRGs indicated that RORA, EAF1, NFYB, miR-4703-3p, and miR-6073 may play a role in the development of ASD. In addition, we found piperaquine may have the potential to be a promising drug for the treatment of ASD. Overall, we constructed a diagnostic model of childhood ASD, which could contribute to the precision diagnosis and timely treatment of childhood ASD.