Electronic medical records are enabling researchers to connect the dots between genes and disease. A paper appearing in Nature Biotechnology from researchers at Vanderbilt University School of Medicine and elsewhere shows that such records can be used to replicate findings from genome-wide association studies — they replicated 51 of the 77 prior associations they examined — and uncover novel associations.
To do this, the researchers took what a phenome-wide association study, or PheWAS, approach. As Carl Zimmer writes in the New York Times, this scheme works by "by turning genome-wide associations on their head." Instead of starting with a condition and looking for a common gene variant, the Vanderbilt-led team started with gene variants and went looking for conditions.
Drawing upon data from the eMERGE Network, the researchers looked for associations between some 3,150 SNPs and 1,360 phenotypes — such as hypertension, obesity, atrial fibrillation, and more — in more than 13,800 people of European descent. They also uncovered more than 60 new links between genes and disease, including an association between IRF4 rs12203592, which has previously been tied to hair and eye color, to actinic keratosis.
"It's a phenomenal proof of concept," Robert Green from Harvard Medical School tells Zimmer.