The Children's Hospital of Philadelphia has received a little over $800,000 from the National Institutes of Health that it will use to mine the electronic medical records of more than 40,000 children under the auspices of the Electronic Medical Records and Genomics, or eMERGE, network.
The funds are administered by the National Human Genome Research Institute and the Eunice Kennedy Shriver National Institute of Child Health and Human Development and are drawn from a $1.6 million pot of money set aside for three teams to conduct three-year pediatric studies as part of the second phase of the network. The total amount awarded for the second phase of eMERGE is $25 million (BI 8/19/2011).
Another $800,000 award under the pediatric eMERGE program went to Cincinnati Children's Hospital Medical Center and Boston Children's Hospital.
NHGRI issued a request for applications for eMERGE pediatric research sites last July, seeking investigators who would conduct comprehensive research on incorporating genomic findings into clinical care.
Projects selected to receive funding were expected to develop innovative approaches to identifying children with high genetic risks or pharmacogenetically important variants; define testing approaches for the return of clinically relevant genetic search results; expand the electronic phenotyping library; and improve analytical methods and tools for use with genotyping platforms, according to the RFA.
CHOP has established a pediatric biorepository that holds samples from 40,000 children, from birth to the age of 21, who will be part of the study. These samples have been genotyped on either Illumina Infinium 550HH, 610Q, or 660Q or Affymetrix 6.0 microarrays.
According to CHOP's grant abstract, researchers at its Center for Applied Genomics will mine the children's EMRs for disease phenotypes and environmental exposure data in up to 40 disease areas and will then link this information with the GWAS data in order to "establish a phenotype/genotype database for future clinical development with other eMERGE sites."
Hakon Hakonarson, an associate professor of pediatrics at the University of Pennsylvania School of Medicine, told BioInform this week that CHOP researchers will be working with a subset of the standard eMERGE disease areas because some of the phenotypes are linked to diseases that typically occur in adults, such as glaucoma and coronary artery disease.
The CHOP team plans to explore phenotypes such as obesity and asthma among others, he said.
Additionally, the researchers plan to contribute information to the eMERGE network about childhood-specific diseases and developmental traits that aren't possible in adult studies, he said.
The researchers also plan to use the EMR data to "determine pharmacogenetic response profiles, both efficacy and adverse events, and search for polymorphisms impacting variation in response to commonly used drugs in the existing pediatric dataset," the abstract states.
Additionally, the CHOP team said it will expand its CLIA- and CAP-certified array analysis workflow "to enable future sharing of genetic/genomic data with the study participants."
In addition to depositing data into the National Center for Biotechnology Information's database of genotypes and phenotypes, CHOP is also working with the other eMERGE sites to create a centralized database that will hold the data from the network. This is expected to make it easier for investigators at the different eMERGE sites to study a single phenotype across all disease cohorts, Hakonarson said.
"That is something which is not possible in dbGAP unless you apply [to the resource] and download all of these individual datasets," he explained. "If each [eMERGE] site did that, it [would be] much more complicated than centralizing it."
CHOP's researchers have also written some scripts that are tailored for the group's in-house EMR system, which allow them to delve into their patients' records and extract any "quantifiable information" — such as hemoglobin measures and radiology test results — while leaving out personal identifiers such as patients' names and dates of birth, Hakonarson said.
The researchers are also using a bespoke version of the NIH's Informatics for Integrating Biology and the Bedside (i2b2) infrastructure, which they use to "categorize" traits, he said.
For example, patients with asthma could be grouped under a single node and then a sub-node could include information such as age, sex, ethnicity, allergy test results, lung function results, medications, and environmental information for the patients, he said.
A researcher could then use this information to define an asthma phenotype to meet specific study requirements. For example, the researcher could select children who are between 5 and12 years of age who have reduced lung function and take steroid medication, and then "go into [the] database resource and call up all the diverse phenotypes for the cohort ... pick a control cohort that matches in terms of age and sex and ... and run [the GWAS] study," he said.
A third goal for CHOP's funding is to establish rules and guidelines that will govern the CAG biorepository and databases and generate informed consent procedures that "optimize existing data and sample use for research and foster clinical utility of the data," the researchers said.
To be included in CHOP's study, children aged seven and older had to assent to participate and their parents were required to provide consent, Hakonarson said.
When the children in the study turn 18, they will have to re-consent to their data being used in the study, he said.