Research
Driving methods and research
Our diverse group of biostatisticians, data scientists, and health economists is comprised of faculty members and graduate student apprentices primarily from the Department of Biostatistics and Informatics in the Colorado School of Public Health at the CU Anschutz Medical Campus, who work to stay on the cutting edge of research and analytics.
We are passionate about data and we work to transform how it is analyzed by conducting our own original research.
Realizing the potential of metabolomics
Professors Katerina Kechris, PhD and Debashis Ghosh, PhD of CIDA were awarded a U01 from the NIH Metabolomics Common Fund Program, which was developed to realize the potential of metabolomics to inform basic, translational, and clinical research. Metabolomics refers to the profiling of the complete repertoire of small molecules—or metabolites—in an individual and provides new opportunities to develop precision medicine. However, the analysis of this repertoire can be challenging due to the size and complexity of metabolomics data resulting from new technologies. The awarded project is entitled "Addressing Sparsity in Metabolomics Data Analysis" and is focused on the development of metabolomic data analysis and interpretation tools to help maximize the potential of metabolomics to make new discoveries. Several Biostatistics and Informatics students and post-doctoral fellows are making progress in methods for metabolomics data reproducibility, normalization, imputation, and biomarker detection, in addition to creating a software database.
Understanding patients who undergo lung transplants
Ryan Peterson, PhD of CIDA helped to write a commercial grant that aims to increase understanding about patients who undergo lung transplants. Lung transplants are serious and rare procedures that have many challenges, one of which is that their prognosis is highly dependent on whether the transplanted lungs can remain stable; if not, the patient may develop Chronic Lung Allograft Dysfunction (CLAD). Unfortunately, it can be difficult and invasive to measure lung allograft stability in ways that ultimately will benefit the patient, as current methods are prone to diagnostic errors that could lead to unneeded expensive and invasive treatments. Peterson’s recently accepted grant will allow him to investigate a new potential biomarker that could produce accurate and timely information about lung allograft stability with a simple blood test. In addition to designing the statistical analysis plan for this study, Peterson will be spearheading the data analysis on this project.
Designing modern clinical trials
Alex Kaizer, PhD of CIDA, has been working on trial and study design. The design of modern trials has grown increasingly complicated with limited assistance provided to appreciate the trade-offs of certain study designs. In the past year, Kaizer has worked on a paper reviewing the use of master protocols in oncology trials with unique considerations for how to control statistical properties and the consequences of certain decisions. In that same vein, he has developed a novel approach with statistical collaborators at the University of Minnesota and Cleveland Clinic for calibrating study designs to facilitate robust, appropriate trials and corresponding statistical analyses. For an unrelated approach, Kaizer also published research on the use of biomarkers to evaluate compliance in a randomized trial across multiple doses and how it can be used to determine the potential causal effect of real-life policy decisions based on the results from clinical trials.
Lung imaging of the future
Nichole Carlson, PhD of CIDA, and Sarah Ryan, PhD Candidate of CIDA, have teamed up with exceptional collaborators Dr. Tasha Fingerlin and Dr. Lisa Maier at National Jewish Health, the nation’s leading hospital for respiratory care, to study a rare interstitial lung disease called pulmonary sarcoidosis. Their goal, awarded with grant funding through the NIH, is to develop reproducible radiographic phenotypes of pulmonary sarcoidosis from computed tomography scans of the lung, and integrate radiographic data with clinical data, genetic variants and transcriptional signatures, which would redefine sarcoidosis biomarkers. They have already seen great success in identifying a strong association between lung function and radiographic features, with ongoing research projects to identify new radiographic phenotypes for sarcoidosis. Further, the statistical techniques being developed extend to lung diseases beyond sarcoidosis, which allow researchers to understand and track disease progression in a sensitive and objective fashion.