In this, our Education and Outreach Core, we plan to sustain and expand the multiple efforts begun in our first cycle to broadly contribute to the """"""""national computational infrastructure"""""""" by (1) growing the base of scientists adept at seeing across the horizon to create innovative intellectual approaches and methodological solutions to our national health care agenda and (2) to broadly enable the community to effectively use our software platform together with the I2b2 discovery model to advance research using existing clinical data and biological byproducts. Our contribution to the next generation of computational scientists begins with our intensive, formally organized Summer Institute In Bioinformatics and Integrative Genomics. Each summer we accept 12-16 undergraduate college students with demonstrated Interest in either bioinformatics or integrative genomics for an intensive immersion experience in the medical applications of these skill sets when combined as an interdiscipline. Our goal is to reach students at the formative stage of their career planning and by exposing them to the potential of a career applying hard science skills (and innovative thinking) to medical problems. Instead of pure didactic information transferral, we ask physicians and researchers already working in this space to give high level tutorials regarding """"""""their"""""""" disease and how bioinformatics had made a difference in the understand and treatment of that disease. It is unbelievably powerful for students to hear about the genetics of deafness, hypertension, diabetes, Huntington's Disease, lymphoma, asthma, rheumatoid arthritis, major depression and bipolar disease and the advancements in therapy made possible by a disciplined analysis of the volumes of data emerging from well designed genetic and genomic studies. Students are simultaneously engaged in a mentored research project with a mentor carefully chosen to match their personal interests and the goals of the program. We have graduated over 50 students, including 29 women and 25 students from under represented minorities. 28 of the 32 College grads are now in graduate school and most are woridng in the bioinformatics or related concentrations. We will put 60 more students through this program over the next four years and anticipate seeing many of them emerge from the pipeline ready and eager to carry the flag. As to our second community, those who use our software platform for translational research, we plan to continue with our support and outreach efforts to ensure the most effective use thereof. As discussed elsewhere, the unanticipated advent ofthe NIH-funded Centers for Translational Science (CTSAs) and the mandate to develop informatics systems to better organize and use clinical data created an unexpected eariy demand for the I2b2 product. We were neither funded for nor prepared to provide support services to those wishing to implement our eariy releases but recognized even so the importance of building a community that could leverage their Individual capabilities and collaborate to serve the whole. We formed and now support an Academic Users'Group that is 115 strong and growing. We convene and will continue to do so, biannual business and working group meetings ofthis group at which we provide updates on what i2b2 is doing, both as regards novel development and for solving issues raised by the membership as they install and use the software. More recently, we have users sufficiently advanced that we request presentations on use cases and use the ensuing dialogue to Identify what the community most wants to see from us. The networking among them, both at meetings and Independently thereafter as a result of knowing who Is doing what with what, etc., has and should continue to greatly leverage existing capabilities, especially when these are limited. As is apparent In the letters of support from this group, the collaborative support afforded by the network has been an essential component to success for many of the CTSAs. We anticipate that the AUG will continue to play a critical role and propose in this next period to ramp up our active support In several ways. Based on the volume of traffic seeking answers to basic implementation questions, we will develop a series of web-based tutorials that will be available on-line and will span the distance from someone shopping for basic insight to what """"""""i2b2"""""""" is to those actually interested in using repurposed clinical care data for genome-phenome association work. The AUG will actively collaborate and participate In this effort. We are especially pleased to propose continuation and expansion of our efforts with the Natural language processing community, which as a group had long been stymied by the dearth of actual clinical records on which to practice their art. Thanks to one of our bright young stars. Dr. Ozlem Uzuner, and the willingness of Partners HealthCare to share a corpus of deidentified clinical notes, i2b2 has hosted three Shared Task in Natural Language Processing for Clinical Data and Workshops. At least a couple dozen teams have joined each year to compete on problems of smoking status, de-identification capability, obesity definition, identification of co-morbidities, and most recently medication status. Papers and many new tools have emerged from this effort, which, as certified by letters from the international community (Research Plan Section 1.6.2), is one they hope will continue. We would like to expand this activity to include other groups In the planning and execution and to add data from other Institutions to the corpus now freely available on our website. Given the growing need for sophisticated NLP tools by the CTSA and other academic health centers, we would like to work with the community to offer an advanced training program and use the I2b2 platform and tools suite as the substrate for this. We will of course also continue to sponsor symposia and other national fora designed to """"""""widen the use of Electronic Health Record data for discovery research"""""""" and anticipate that increasingly we will be able to widen the repertoire of investigators with compelling stories to tell.

National Institute of Health (NIH)
National Library of Medicine (NLM)
Specialized Center--Cooperative Agreements (U54)
Project #
Application #
Study Section
Special Emphasis Panel (ZRG1-BST-K)
Project Start
Project End
Budget Start
Budget End
Support Year
Fiscal Year
Total Cost
Indirect Cost
Brigham and Women's Hospital
United States
Zip Code
Yu, Zhi; Kim, Seoyoung C; Vanni, Kathleen et al. (2018) Association between inflammation and systolic blood pressure in RA compared to patients without RA. Arthritis Res Ther 20:107
Agniel, Denis; Kohane, Isaac S; Weber, Griffin M (2018) Biases in electronic health record data due to processes within the healthcare system: retrospective observational study. BMJ 361:k1479
Yu, Sheng; Ma, Yumeng; Gronsbell, Jessica et al. (2018) Enabling phenotypic big data with PheNorm. J Am Med Inform Assoc 25:54-60
Luo, Yuan (2017) Recurrent neural networks for classifying relations in clinical notes. J Biomed Inform 72:85-95
Nanba, Kazutaka; Vaidya, Anand; Williams, Gordon H et al. (2017) Age-Related Autonomous Aldosteronism. Circulation 136:347-355
Kim, Youngjun; Riloff, Ellen; Meystre, Stéphane M (2017) Exploiting Unlabeled Texts with Clustering-based Instance Selection for Medical Relation Classification. AMIA Annu Symp Proc 2017:1060-1069
Bigdeli, T B; Ripke, S; Peterson, R E et al. (2017) Genetic effects influencing risk for major depressive disorder in China and Europe. Transl Psychiatry 7:e1074
Hundemer, Gregory L; Baudrand, Rene; Brown, Jenifer M et al. (2017) Renin Phenotypes Characterize Vascular Disease, Autonomous Aldosteronism, and Mineralocorticoid Receptor Activity. J Clin Endocrinol Metab 102:1835-1843
Luo, Yuan; Wang, Fei; Szolovits, Peter (2017) Tensor factorization toward precision medicine. Brief Bioinform 18:511-514
Luo, Yuan; Uzuner, Özlem; Szolovits, Peter (2017) Bridging semantics and syntax with graph algorithms-state-of-the-art of extracting biomedical relations. Brief Bioinform 18:160-178

Showing the most recent 10 out of 312 publications