EHR-Based Genomic Discovery

As part of the Electronic Medical Records and Genomics (eMERGE) Network, the research team in Dr. Kullo's Cardiovascular Biomarkers Lab is conducting electronic health record (EHR)-based genetic studies.

The network is funded by the National Human Genome Research Institute (NHGRI) to develop and implement approaches for leveraging biorepositories with EHR systems for large-scale genomic research, including genome-wide association studies (GWAS), sequencing and structural variation.

The five initial eMERGE sites were:

  • Group Health Cooperative (with the University of Washington)
  • Marshfield Clinic
  • Mayo Clinic
  • Northwestern University
  • Vanderbilt University
Related publications

Mayo eMERGE phase I project

The Mayo Clinic eMERGE I project aimed to identify genetic loci associated with peripheral arterial disease and red blood cell traits, including hemoglobin, hematocrit, red blood cell count, mean corpuscular volume, mean corpuscular hemoglobin and mean corpuscular hemoglobin concentration.

Phenotype characteristics and covariates relevant to statistical genetic analyses were derived using EHR-based algorithms, including diagnosis and procedure codes, laboratory data, medication use, and natural language processing of unstructured text. An ethical, legal and social issues (ELSI) component was included and focused in particular on community interaction and return of results.

Mayo eMERGE phase II project

In phase II of eMERGE (July 2011 to June 2015), the research infrastructure established in eMERGE-I was leveraged to identify common genetic variants that influence medically important phenotypes.

The Mayo eMERGE-II cohort (6,916 people) includes the 3,769 eMERGE-I patients and an additional 3,147 people, the majority (90 percent) of whom were genotyped on the same Illumina 660W platform. A major focus in phase II was to translate recent GWAS findings into clinical practice.

Related publications

Mayo eMERGE phase III project

The eMERGE III study is an exciting collaboration between Mayo Clinic and Mountain Park Health Center in Arizona that's funded by the NHGRI and the Mayo Clinic Center for Individualized Medicine.

Mayo Clinic's eMERGE III study investigated the return of genetic test results relevant to hypercholesterolemia and colon polyps.

As part of the study, genetic testing for hypercholesterolemia and colon polyps will be performed at a National Institutes of Health laboratory. A team of expert Mayo Clinic investigators will look for the presence of genetic abnormalities causing hypercholesterolemia or colon polyps.

The genetic results will be entered into the EHR. Patients will receive genetic counseling and support from Mayo Clinic physicians and genetic counselors.

In eMERGE III, the participating sites were expanded and include:

  • Children's Hospital of Philadelphia
  • Cincinnati Children's Hospital Medical Center
  • Columbia University
  • Geisinger Health System
  • Group Health Research Institute/University of Washington
  • Mayo Clinic
  • Northwestern University
  • Seattle Brigham and Women's Hospital
  • Vanderbilt University School of Medicine

Study 1: EHR-based algorithms to ascertain vascular disease phenotypes

With the decreasing cost of genotyping and the increasing availability of genomic data from consortia, selection of phenotypes of interest becomes a critical part of phenotype-genotype association studies.

As an important node of the eMERGE Network, our Cardiovascular Biomarkers Lab designed several EHR-based phenotyping algorithms that can be found on the Phenotype KnowledgeBase (PheKB) website. PheKB is a collaborative environment enabling cross-site collaboration for algorithm development, validation and sharing for reuse with confidence.

Related publications

Study 2: Genome-wide association studies

We have performed EHR-based genome-wide association studies of peripheral arterial disease and red blood cell indices.

Related publications

Study 3: Phenome-wide association studies (PheWAS)

PheWAS is a unique tool to look at traits associated with a particular therapy, enabling discovery of pleiotropic effects of genes.

This strategy mines the full potential of the EHR for genome-phenome associations, rapid and efficient validation of known associations between genes of interest, and interindividual variation in traits.

Our Cardiovascular Biomarkers Lab previously detected signals of natural selection in PCSK9, and recent animal and genetic studies implicating PCSK9 in the regulation of blood pressure and triglyceride-rich lipoprotein metabolism indicate that genetic variants in PCSK9 may have pleiotropic effects.

We are conducting a comprehensive investigation of pleiotropic effects of PCSK9.

Related publications