This career transition award will provide the applicant with protected research time to develop an independent research career in biomedical informatics. Dr. Starkey received his M.D. in 2003 from UT Southwestern, completed the NLM Biomedical Informatics Training Program post-doctoral fellowship in 2011 and earned a Ph.D. focusing on clinical informatics from the Clinical Sciences Degree Program at the University of Texas Medical Branch (UTMB) at Galveston in 2012. Dr. Starkey accepted a faculty position at the Institute for Translational Sciences at UTMB that is currently supporting his transition to independent investigator. Dr. Starkey will utilize the career transition award for novel applications of advanced machine learning methods to create predictive models of disease. Predictive models of disease allow for the stratification of risk and prevention of disease and have been successfully implemented in the assessment of cardiovascular disease risk. Predictive models of chronic kidney disease (CKD) are an active area of research since CKD is a risk factor for all-cause mortality, cardiovascular death and end-stage renal disease. However, there is not a single predictive model of CKD created for application to Hispanics and the external validity of existing models is poor. The objective of this project is to create predictive models of chronic kidney disease in a Hispanic population and quantify biomarkers of chronic kidney disease using multiple reaction monitoring (MRM) proteomics in this minority population. MRM proteomics will utilize the UTMB Novel Methodologies core and its array of resources to create an inter-institutional collaboration. Heterogeneous ensemble machine learning methods will be used to create the predictive model of chronic kidney disease in the Cameron County Hispanic Cohort (CCHC). Established in 2004, the CCHC is a random population sample in Brownsville, Texas created to evaluate the determinants of health in a US/Mexico border population that is primarily of Hispanic ethnicity. The CCHC has an extremely high prevalence of obesity and diabetes at 49.7% and 30.3%, respectively. Both obesity and diabetes are independent risk factors for the development of CKD and represent significant health disparities in the Hispanic population. Archived serum samples of CCHC participants and the CCHC database will be utilized to complete the following aims: 1) Create predictive models of CKD in a Hispanic population 2) Refine the predictive models by including clinical laboratory data and a selective reaction monitoring mass spectrometry panel of biomarkers. At the completion of this project, predictive models of CKD applicable to a Hispanic population will be created and MRM proteomics will demonstrate utility of biomarkers identified in other populations. The application of the results will be used to create clinical tools for CKD risk and guide future studies in this Hispanic population with health disparities. Importantly, it will provide a career transition for D. Starkey to demonstrate the knowledge gained during his training that results in impactful research and publications.

Public Health Relevance

The proposed research is relevant to public health because it allows stratification of chronic kidney disease risk and quantifies biomarkers of chronic kidney disease in the Hispanic minority whom has significant health disparities with regard to kidney disease. The translation of selective reaction monitoring of biomarkers and development of heterogeneous ensemble machine learning methods for application to the prediction of chronic kidney disease in a minority population supports the NLM mission.

National Institute of Health (NIH)
National Library of Medicine (NLM)
Career Transition Award (K22)
Project #
Application #
Study Section
Biomedical Library and Informatics Review Committee (BLR)
Program Officer
Sim, Hua-Chuan
Project Start
Project End
Budget Start
Budget End
Support Year
Fiscal Year
Total Cost
Indirect Cost
University of Texas Medical Br Galveston
Schools of Medicine
United States
Zip Code