The Vanderbilt Genome-Electronic Records project was one of five sites in the first phase of the Electronic Medical Records and Genomics (eMERGE-l) network. The VGER team contributed importantly to progress in multiple areas across eMERGE-l, including developing and deploying algorithms for phenotypes at Vanderbilt and across the network;developing and managing the genotype quality control pipeline for the network;discovering new genotype-phenotype relations in Vanderbilt and cross-network datasets;actively participating in the community consultation and return of results initiatives;developing methods to resolve tensions between data access and individual privacy;and developing new software tools for the field, including for de-identification and for the phenome-wide association study paradigm (""""""""PheWAS""""""""). Vanderbilt has identified Personalized Healthcare as a key strategic priority for investment in basic discovery, translation, and implementation across the institution;resources developed at Vanderbilt (directed by VGER team members) include StarChart, an electronic medical record (EMR) system with comprehensive patient-specific clinical decision support functionality that includes data on >1.7 million patients;BioVU, the DNA repository that links >100,000 DNA samples to a de-identified image of the EMR;and the PREDICT project which has created a framework for evaluating genotype-phenotype relations and is depositing clinically actionable genotypes into the EMR. BioVU currently includes 5,186 samples with GWAS data, projected to >16,500 by fall 2011. The present proposal to participate as a site in eMERGE-ll builds on this record of accomplishment and on institutional investments. We propose 4 specific aims that will be accomplished by collaborations among scientists with expertise in diverse disciplines (clinical medicine, basic science, genomics, statistics, informatics, privacy science, and ethics) at our site and across the network: (1) to accelerate development and validation of algorithms for phenotype extraction from EMRs. (2) To exploit the results of GWAS in BioVU and other datasets to identify combinations of genotypes highly predictive of disease or drug response outcomes. (3) To engage patients as we implement prospective clinical genotyping in PREDICT. (4) To develop new tools to maximize our ability to effectively share genomic information and ensure patient confidentiality. We subscribe to a vision of Personalized Medicine in which genomic and other patient-specific information drives healthcare, and VGER and eMERGE-ll represent important steps in that direction.
Descriptions of how genetic variation determines variability in clinically important conditions like disease susceptibility or drug responses represent the first fruits of the Human Genome Project. A challenge - that this proposal addresses - is how to analyze and use this torrent of information to improve human health. Our proposal to join the eMERGE-ll network addresses this challenge by identifying genetic variants important for human health and beginning to use these in a systems approach to personalized healthcare that is robust and scalable in the face of the escalating volume and complexity of clinically relevant data.
Showing the most recent 10 out of 144 publications