Blood-based DNA methylation markers for autism spectrum disorder identification using machine learning.

Scientists studied blood samples from 100 children (52 with autism, 48 without) to see if they could find genetic patterns that might help identify autism. They found 138 differences in how genes are 'switched on or off' between children with and without autism. Using computer programs, they could correctly identify autism in about 7 out of 10 children based on these blood patterns. While promising, this early research needs much more testing before it could be used as a diagnostic tool.

This exploratory study investigated whether DNA methylation patterns in blood samples could help identify autism spectrum disorder (ASD) using machine learning. Researchers analyzed blood samples from 52 children with ASD and 48 typically developing children, identifying 138 DNA methylation differences between groups. Using machine learning algorithms, they selected 11 key genetic markers and built classification models that achieved 70-75% accuracy in distinguishing children with ASD. The identified markers were enriched in pathways related to cell adhesion and immune function.

While this represents a novel approach to ASD biomarker development, the study was limited by small sample size and moderate classification accuracy, requiring further validation in larger populations.

While not ready for clinical use, this research demonstrates proof-of-concept for blood-based epigenetic biomarkers in ASD. The moderate accuracy and small sample size require significant improvement through larger validation studies before potential clinical translation. Current findings support continued research into epigenetic mechanisms underlying ASD.

Small sample size (100 participants), moderate classification accuracy (70-75%), cross-sectional design, single dataset analysis, and lack of independent validation cohort. Study acknowledges exploratory nature and methodological limitations that prevent clinical application.

Autism spectrum disorder (ASD) is a complex neurodevelopmental disorder lacking objective biomarkers for early diagnosis. DNA methylation is a promising epigenetic marker, and machine learning offers a data-driven classification approach. However, few studies have examined whole-blood, genome-wide DNA methylation profiles for ASD diagnosis in school-aged children. We analyzed genome-wide DNA methylation data from GEO dataset GSE113967, including 52 children with ASD and 48 typically developing (TD) controls.

Differentially methylated positions (DMPs) were identified, and feature selection was performed using support vector machine-recursive feature elimination with cross-validation (SVM-RFECV). Classification models were developed using random forest (RF), extreme gradient boosting (XGBoost), and decision tree (DT) classifiers. A nomogram visualized feature contributions. A total of 138 DMPs differentiated ASD from TD children.

Eleven CpG sites selected by SVM-RFECV formed the basis for model construction. RF and XGBoost achieved the highest accuracy (75%), with DT reaching 70%. Functional annotation indicated enrichment in cell adhesion and immune-related pathways. This exploratory study demonstrates the feasibility of integrating peripheral blood DNA methylation data with machine learning to distinguish children with ASD.

While limited by sample size and moderate accuracy, this study provides methodological insights into the feasibility of integrating epigenetic and computational approaches for ASD-related biomarker exploration.