Academics dig into EMRs to discover genetic causes of disease

In the 13 years since the publication of the rough draft of the human genetic code, scientists have run thousands of genome-wide association studies to find links between DNA and disease. The work has delivered some insights, but also shown the method has limitations. Now, a new approach could avoid these flaws.

The new method uses electronic medical records (EMRs) to reverse the genome-wide association model. In genome-wide studies, researchers search for common mutations in the DNA of people with the same condition, like diabetes or Alzheimer's disease. The new approach, called a phenome-wide association study, flips this model. Scientists start with a gene mutation and then scour EMRs for possible links between the variant and disease.

A test of the model published in Nature Biotechnology by researchers from the Vanderbilt University Medical Center and other academic institutions has excited geneticists. "It's a phenomenal proof of concept. Warts and all, we can use [EMRs] to do science," Harvard Medical School geneticist Robert Green told The New York Times. Researchers validated the model by replicating results from genome-wide studies, and then used it to identify 63 new links between genes and diseases like skin cancer and anemia.

Joshua Denny, Vanderbilt researcher

The validation of the EMR model gives drug developers a new tool. Identification of the genetic causes of disease could give developers new, potentially-druggable targets to pursue. The Vanderbilt team also want to spot potential drug side effects using the model, and identify new uses for existing compounds. "If you have a drug that targets a certain gene, you can understand what range of diseases you can use that drug to treat," Vanderbilt biomedical informatics researcher Joshua Denny said.

- read the NYT article (sub. req.)
- here's the Nature abstract
- check out the press release