My work has generally involved application of innovative biostatistical approaches to big data sources (previous name Duarte). For instance, we performed an analysis of Medicare claims data linked to a national cancer registry (SEER, Surveillance, Epidemiology, and End Results) to discover that individuals with HHT (Hereditary Hemorrhagic Telangiectasia), a rare genetic disorder involving haploinsufficiency of the protein endoglin, showed improved cancer outcomes compared to non-HHT patients (1) but no association with cancer incidence (2). We have also used a novel network analysis to analyze the association between cancer and cardiovascular disease using the UK Biobank hospital claims data (3), and we have developed and validated prediction models for end of life planning (4), cancer prediction (5), and pharmacogenomic response (6).
- Duarte CW, Murray K, Lucas FL, Fairfield K, Miller H, Brooks P, et al. Improved survival outcomes in cancer patients with hereditary hemorrhagic telangiectasia. Cancer Epidemiol Biomarkers Prev. 2014;23(1):117-25.
- Duarte CW, Black AW, Lucas FL, Vary CP. Cancer incidence in patients with hereditary hemorrhagic telangiectasia. J Cancer Res Clin Oncol. 2016.
- Duarte CW, Lindner V, Francis S, Schoormans D. Visualization of cancer and cardiovascular disease co-occurrence using network methods. Journal of Clinical Oncology/Clinical Cancer Informatics. 2017.
- Duarte CW, Black A, Murray K, Haskins A, Lucas L, Hallen S, et al. Validation of the Patient-Reported Outcome Mortality Prediction Tool (PROMPT). Journal of Pain and Symptom Management. 2015.
- Duarte CW, Willey CD, Zhi D, Cui X, Harris JJ, Vaughan LK, et al. Expression signature of IFN/STAT1 signaling genes predicts poor survival outcome in glioblastoma multiforme in a subtype-specific manner. PLoS One. 2012;7(1):e29653.
- Cosgun E, Limdi NA, Duarte CW. High-dimensional pharmacogenetic prediction of a continuous trait using machine learning techniques with application to warfarin dose prediction in African Americans. Bioinformatics. 2011;27(10):1384-9.
Future Directions
My future research interests involve continuing to use epidemiological and multiomic approaches to discover molecular mechanisms of disease. The translational potential of this work includes discovery of biomarkers and new therapeutic targets. In my collaborative work, it is my goal to discover the translational relevance of key genes and pathways through secondary analysis of ‘omic data sources. A new research interest of mine is developing prediction models using machine learning or penalized regression approaches of genomic network modules for robust prediction of disease outcomes or treatment response.