By Kelly Liu
A study led by Fabric Genomics and Rady Children’s Institute for Genomic Medicine has demonstrated that artificial intelligence can lead to the accurate and rapid clinical diagnoses of rare diseases in critically ill newborns. Clinical interpretation of genetic variants in patient phenotype is becoming the largest component of high cost and time for genome-based diagnosis of rare genetic diseases, and the hope is that AI may be able to expedite this process. The new AI-based genome interpretation tool, Fabric GEM, has shown promising results.
About 7 million infants are born with serious genetic disorders each year worldwide, and it can take days or weeks to diagnose the illness from the genome. Arriving at a diagnosis within the first 24 to 48 hours after birth is critical for the most effective treatment and the best chance to improve the patients’ conditions. To test GEM’s ability to find the DNA errors that lead to these diseases, scientists used GEM to run whole genomes from 179 previously diagnosed pediatric cases from six leading genomic centers and hospitals. GEM detected the causative gene as one of its top two candidates 92% of the time. Existing tools have accomplished the same task less than 60% of the time. Additionally, Fabric GEM ranked specific diseases and conditions associated with these genes, acting as further assistance to clinicians in the ultimate diagnosis of each case. The GEM AI algorithm cross-references a vast and ever-growing body of large databases of genomic sequences from diverse populations, clinical disease information and other sources of medical and scientific data, combining this information with the patient’s genome sequence and medical records to make the most accurate predictions possible.
Existing technologies mainly identify small genomic variants such as single DNA letter changes or small strings of insertions or deletions. GEM is more advanced in that it can also find “structural variants” as causes of disease that are larger and often more complex. It is estimated that structural variants are behind 10% to 20% of genetic diseases, making GEM a much more comprehensive and reliable tool for applications in diagnosing patients in places such as the NICU (neonatal intensive care unit).
Furthermore, GEM can be coupled with a natural language processing tool, Clinithink’s CLiX focus, which scans reams of doctors’ notes for the clinical presentations of the patient’s disease. This removes the need for physicians to manually review and summarize note contents as a part of the diagnostic process, forwarding the medical record search and increasing efficiency in speed and scalability.
“Finally, clinicians do not have to sacrifice accuracy for speed when faced with a possible rare disease diagnosis in a critical setting like the NICU where time is of the essence,” said Martin Reese, PhD., CEO of Fabric Genomics. Fabric GEM makes genome sequencing more cost-effective and quick, and is a major innovation that allows clinicians to better provide an explanation for causes of an illness, improve disease management, and, hopefully, lead to recovery.
Resources:
De La Vega, F. M., Chowdhury, S., Moore, B., Frise, E., McCarthy, J., Hernandez, E. J., … Kingsmore, S. F. (2021). Retrieved from https://genomemedicine.biomedcentral.com/articles/10.1186/s13073-021-00965-0
Kiefer, J. (2021). Ai quickly identifies genetic causes of disease in newborns. Retrieved from https://attheu.utah.edu/facultystaff/ai-technology-identifies-genetic-causes-of-serious-disease/
Reese, E. (2021). Retrieved from https://fabricgenomics.com/2021/10/benchmark-genome-study-demonstrates-accuracy-of-artificial-intelligence-in-rapidly-diagnosing-rare-diseases-in-critically-ill-patients/
Comments