New AI tool improves discovery of genes involved in neurodevelopmental conditions

New AI tool improves discovery of genes involved in neurodevelopmental conditions

Researchers have developed an artificial intelligence (AI) approach that accelerates the identification of genes that contribute to neurodevelopmental conditions such as autism spectrum disorder, epilepsy and developmental delay. This new powerful computational tool can help fully characterize the genetic landscape of neurodevelopmental disorders. This is the key to accurate molecular diagnosis, elucidation mechanism and development of targeted therapies. The study appeared in theAmerican Journal of Human Genetics.

"Although researchers have made important advances identifying various genes associated with neurodevelopmental disorders, many patients with these disorders still do not receive a genetic diagnosis, indicating that there are many more genes waiting to be discovered," said first and co-correcting author Dr.

To discover new genes associated with a disease, researchers sequence the genomes of many people with the disorders and compare them to the genomes of people without the disorders.

We have taken a complementary approach. We used AI to find patterns between genes already linked to neurodevelopmental diseases and predict additional genes that are also involved in these disorders. “

Dr. Ryan S. Dhindsa, assistant professor of pathology and immunology, Baylor College of Medicine

The researchers looked for patterns of gene expression measured at the single-cell level from the developing human brain. "We found that AI models trained solely on this expression data can robustly predict genes involved in autism spectrum disorders, developmental delay and epilepsy. However, we wanted to take this work a step further," said Dhindsa.

To further improve the models, the team included more than 300 additional biological features, including measures of how intolerant genes are to mutations, whether they interact with other known disease-associated genes, and their functional roles in various biological pathways.

“These models have exceptionally high predictive value,” Dhindsa said. "The top genes were up to two-fold or six-fold, depending on the type of inheritance, being enriched for the high-confidence disorder risk genes rather than the genetic intolerance metrics alone. Additionally, some top-ranking genes were 45 to 500 times more likely to be supported by the literature lower than lower-ranking genes."

“We see these models as analytical tools that can validate genes that emerge from sequencing studies but do not yet have enough statistical evidence to be involved in the conditions for neurodevelopment,” Dhindsa said. “We hope our models will accelerate gene discovery and patient diagnoses, and future studies will evaluate this possibility.”

Blake A. Weido, Justin S. Dhindsa, Arya J. Shetty, Chloe F. Sands, Slavé Petrovski, Dimitrios Vitsios, and joint author Anthony W. Zoghbi contributed to this work. The authors are affiliated with one or more of the following institutions: Baylor College of Medicine, Jan and Dan Duncan Neurological Research Institute at Texas Children's Hospital, AstraZeneca, and the University of Melbourne.

This work is supported by grants from NIH Ninds (F32 NS127854), NIH (DP5 OD036131), a longevity grant from the Norn Group, the Hevolution Foundation, the Rosenkranz Foundation, and Grant K23MH121669.

Sources: