New Research Highlights Sex Differences in Genetic Prediction of Autism Spectrum Disorder
Sex-Specific Genetic Factors Influence Autism Prediction
Recent research has shed light on how biological sex may affect the accuracy of genetic predictions for autism spectrum disorder (ASD). ASD affects about 1% of the global population, with males diagnosed approximately four times more often than females. While genetic models have been developed to forecast ASD risk, the role of sex in these predictions has not been thoroughly examined.
Study Overview and Methodology
A team of researchers developed an ensemble prediction model that combines polygenic risk scores (PRSs), common genetic variants, and known ASD risk genes. This model was trained and tested using the MSSNG whole genome sequencing dataset, which includes extensive genetic information from individuals with ASD.
Key Findings
The model demonstrated moderate predictive performance, achieving an accuracy of 68%, an area under the receiver operating curve (AUROC) of 0.72, and a recall rate of 77% on the test data. Importantly, the contribution of common genetic variants to ASD prediction was significantly stronger in males than females (p < 0.001), with prediction accuracies of 69% for males versus 66% for females.
Furthermore, a specific genetic region known as the 16p11 locus was found to be particularly predictive for females (p < 0.001). Gene expression analyses revealed that ASD risk genes significant in females were enriched in brain areas such as the primary somatosensory cortex, inferior parietal cortex, and parietal neocortex during fetal development. In contrast, genes associated with ASD risk in males showed enrichment in the dorsolateral prefrontal cortex and anterior cingulate cortex throughout developmental stages from fetal life to adulthood.
Implications for Families and Future Research
These findings emphasize the importance of considering biological sex when using genetic information to assess ASD risk. The differences in brain region gene expression patterns suggest that ASD may manifest through distinct neurodevelopmental pathways in males and females. This insight could inform more personalized approaches to early diagnosis and intervention.
Limitations and Considerations
While the model shows promise, its predictive accuracy remains moderate and should not be used as a standalone diagnostic tool. The study highlights associations rather than direct causation, and further research is necessary to understand the complex interplay between genetics, sex, and ASD development.
For more detailed information, the original study can be accessed here.
Download the images in this article
