The gene mutation causes a loss of function in the HSD17B13 gene, which alters the function of a liver protein, according to a March 21 statement from Geisinger. The researchers found people with two copies of the variant have a lower risk of various types of liver diseases, including a 73 percent lower risk of alcoholic cirrhosis of the liver and 49 percent lower risk of non-alcoholic cirrhosis.
For the study, the researchers analyzed DNA sequencing data and EHR data from 46,544 Geisinger MyCode Community Health Initiative participants, along with data from Dallas-based University of Texas Southwestern’s Dallas Heart Study and the Philadelphia-based Penn Medicine Biobank. They also collected liver tissue samples from Geisinger’s Obesity Institute.
Geisinger’s MyCode is a genomics and precision health program targeting patients across Pennsylvania and New Jersey. Participants donate their DNA through a blood sample for research, and consent to researchers accessing their sequenced DNA data and EHRs to study various clinical risk factors. To date, more than 190,000 participants have enrolled in the project.
The researchers suggested their finding might help future investigators develop medicines that mimic the gene variant to reduce the risk of chronic liver diseases. Regeneron Pharmaceuticals, which runs the Regeneron Genetics Center, plans to collaborate with Alnylam Pharmaceuticals to develop RNAi therapeutics based on the gene target.
“This is exciting news because it means that future therapies that mimic the effect of this genetic change could have the same protective effect against liver disease,” said Tooraj Mirshahi, PhD, one of the researchers on the project and an associate professor in the functional and molecular genomics department at Geisinger.
More articles on data analytics & precision medicine:
EHR data predicts chronic opioid use at safety-net hospital, study suggests
Intermountain Healthcare to build global DNA registry with 23andMe, AncestryDNA data
IBM releases new cloud platform for data science, machine learning
