A framework to infer de novo exonic variants when parental genotypes are missing enhances association studies of autism.

Researchers created a computer program that can identify important genetic changes linked to autism, even when parents' DNA isn't available for comparison. Usually, scientists need DNA from both parents and children to find new genetic mutations. This new method helps researchers study autism genetics more effectively when they only have the child's DNA sample, potentially leading to better understanding of autism's genetic causes.

This study presents a computational framework to identify de novo (new) genetic variants in autism research when parental DNA samples are unavailable. The researchers developed a machine learning classifier that can distinguish between inherited and de novo variants using only the affected individual's genetic data. They created a scoring system to predict the likelihood that a variant is de novo and integrated this into a Random Draw model for gene discovery. The framework enhances statistical power for identifying autism-associated genes while controlling false discovery rates.

The methodology addresses a common limitation in genetic studies where complete family data is not available.

This computational framework could accelerate autism genetic research by enabling analysis of samples lacking parental DNA. May lead to identification of additional autism-associated genes, potentially informing genetic counseling and personalized interventions in the future.

Sample size not reported. Methodology is computational/statistical rather than experimental validation. Limited to exonic variants only. Performance metrics and validation on independent datasets not described in abstract.

Gene-damaging mutations are highly informative for studies seeking to discover genes underlying developmental disorders. Traditionally, these de novo variants are recognized by evaluating high-quality DNA sequence from affected offspring and parents. However, when parental sequence is unavailable, methods are required to infer de novo status and use this inference for association studies. We use data from autism spectrum disorder to illustrate and evaluate methods.

Separating de novo from rare inherited variants is challenging because the latter are far more common. Using a classifier for unbalanced data and variants of known inheritance class, we build an inheritance model and then a de novo score for variants when parental data are missing. Next, we propose a new Random Draw (RD) model to use this score for gene discovery. Built into an existing inferential framework, RD produces a more powerful gene-based association test and controls the false discovery rate.

Codes are available at Github (https://github.com/HaeunM/TADA-RD) and Zenodo (DOI: https://doi.org/10.5281/zenodo.18531769).