Artificial Intelligence Can Rapidly Diagnose Rare Genetic Diseases

By Megan Brooks

October 21, 2021

NEW YORK (Reuters Health) - An artificial intelligence (AI)-based algorithm can quickly diagnose rare genetic diseases in critically ill children with high accuracy, allowing earlier treatment, according to a new report.

"This study is an exciting milestone demonstrating how rapid insights from AI-powered decision support technologies have the potential to significantly improve patient care," Dr. Mark Yandell, professor of human genetics at University of Utah Health, and a founding scientific advisor to Fabric Genomics, said in a joint news release from Fabric Genomics and Rady Children's Institute for Genomic Medicine, in San Diego, California.

Worldwide, about 7 million infants are born with serious genetic disorders each year. Dr. Yandell's group worked with Fabric Genomics to develop the Fabric GEM algorithm, which incorporates AI to find DNA errors that lead to genetic diseases.

The study team assessed the diagnostic performance of the Fabric GEM algorithm by analyzing whole genomes from 179 children previously diagnosed with rare genetic diseases at Rady Children's Hospital and five other medical centers around the world.

The AI-powered algorithm identified the disease-causing variant as one of its top two candidates more than 90% of the time, outperforming existing tools, which accomplished the same task less than 60% of the time, the researchers report in Genome Medicine.

"Fast and definitive genetic diagnosis is essential to providing the right treatment in a timely manner for critically ill newborns," Dr. Stephen Kingsmore, a co-author on the study and the president and CEO of Rady Children's Institute for Genomic Medicine, said in the release.

"Fabric GEM has successfully demonstrated that it can automatically and quickly suggest a very short list of candidate genes for interpretation through whole-genome or whole-exome sequencing," Dr. Kingsmore said.

"We're taking genomics into the clinic (and) diagnosing children in the NICU (neonatal-intensive-care unit) within 24 hours or less and then the doctors can immediately change treatment within 24 hours. That has an immediate impact," Dr. Martin Reese, CEO of Fabric Genomics and a co-author on the paper, noted in a phone interview with Reuters Health.

It's thought, Dr. Reese explained, that 40% of children younger than 100 days old who end up in the NICU have a genetic disease "and this is a perfect diagnostic tool for them."

With currently available drugs for rare diseases and an accurate diagnosis, "40% to 50% of these kids can be treated today." For the remaining 50%, there will be new drug development and new drugs coming up, but making an early and accurate diagnose is key, Dr. Reese said.

He said the algorithm would also be useful in adults with rare genetic diseases who may never get a proper diagnosis.

The collaborative research by Fabric Genomics and Dr. Yandell's team at the Utah Center for Genetic Discovery has been supported by several national agencies, including the National Institutes of Health and American Heart Association, and by the University of Utah's Center for Genomic Medicine.

Several authors on the paper are employees of Fabric Genomics or consult for the company.

SOURCE: Genome Medicine, online October 14, 2021.