Skip to main content
Premium Trial:

Request an Annual Quote

Combination of Sequencing, EHR Data Links Genes to Range of Conditions

NEW YORK – By analyzing DNA sequencing data combined with electronic health record information, researchers have uncovered rare variants that implicate about two dozen genes in conditions like glaucoma and diabetes.

Rare variants may have a greater effect on disease risk, but can be, due to their infrequency, more difficult to study. Using electronic health records and DNA sequencing data collected on nearly 11,000 individuals, researchers from the University of Pennsylvania conducted an exome-by-phenome-wide analysis to uncover gene-phenotype associations.

As they reported in Nature Medicine on Monday, they used this approach to uncover 26 genes associated with a range of phenotypes, a number of which like PPP1R13L, RGS12, and CLIP had not before been linked to those conditions. These and subsequent findings, the researchers said, could provide greater insight into the biology of human disease. 

"We suggest that application of this approach to even larger numbers of individuals will provide the statistical power required to uncover unexplored relationships between rare genetic variation and disease phenotypes," senior author Daniel Rader from the University of Pennsylvania and his colleagues wrote in their paper.

The researchers searched the 10,900 individuals in the Penn Medicine Biobank for carriers of rare, predicted loss-of-function variants, such as frameshift insertions or deletions, stop codon gain or loss, or splice site disruptions. For genes that had at least 25 heterozygous carriers of a predicted loss-of-function variant, the researchers then looked to see whether there was an associated phenotype. In all, 97 genes had exome-by-phenome-wide significant phenotype associations.

Using a separate cohort of 6,432 African-Americans from the Penn Medicine Biobank as well as data from Mount Sinai's BioMe biobank, the Geisinger Health System's DiscovEHR biobank, and the UK Biobank, the researchers replicated 26 of these findings.

Of those, five associations were known and served as a type of positive control for gene-disease associations. For instance, rare, predicted loss-of-function variants in CFTR were significantly associated with cystic fibrosis, which is caused by biallelic CFTR variants. 

The 21 other associations, though, were novel. Rare, predicted loss-of-function variants in PPP1R13L, an inhibitor of p53, were associated with primary open-angle glaucoma; while variants in RGS12 were linked to type 1 diabetes and ones in CLIP with aortic ectasia, which is usually linked to connective tissue disorders.

Functional analyses further supported the roles of some of these genes in the phenotypes at hand. Using the Ocular Tissue Database, the researchers found PPP1R13L is highly expressed in ocular tissues involved in glaucoma. Further, in a mouse model of glaucoma, they found that PPP1R13L expression in the optic nerve head is highest in late-early to moderate stages of the disease, and noted that PPP1R13L inhibition worsens the death of retinal ganglion cells — which are the most affected in glaucoma — after axonal injury. At the same time, in human induced pluripotent stem cells, the researchers found oxidative stress upregulates PPP1R13L expression.

Based on this, they concluded that PPP1R13L is upregulated by oxidative stress and may prevent retinal ganglion cell death from p53 activation and p53-mediated apoptosis in glaucoma, and that PPP1R13L haploinsufficiency contributes to the effects of glaucoma.

While the cohorts the researchers examined were largely of European ancestry, about 20 percent were of African ancestry. They additionally uncovered 16 rare, predicted deleterious single variants specific to African ancestry, none of which were in the GWAS catalog or mentioned in the literature. This suggested to them that larger, more diverse studies are needed to improve the understanding of how variants like these may contribute to disease.