My lab has a major focus on biomarker discovery and validation using multi-omics technologies, with a particular focus on proteomics data. We are generating large datasets of proteomics data for various precision medicine and diagnostic, predictive, and prognostic applications in human disease. The key challenge with these big data is the development of the most accurate predictor or risk stratification models and to define the pathophysiological pathways associated with these biomarkers. The goal of the project is to apply statistical, machine learning or AI approaches to these big data in order to develop the optimal multi-feature predictor or risk stratification models that eventually can be translated into clinical applications to improve patient management and outcome. Systems biology approaches will also be applied to generate novel insights into the pathophysiological pathways underlying disease development or progression or response to therapy. Larger datasets for various human diseases are available or being generated every week that are available for a student to further analyze. The student will be involved in all aspects of data analysis with the goal to develop the best models with regard to prediction and biological function. Prior skills should include knowledge in at least one of the listed computer languages such as R, Python, Matlab or SAS. Some statistical or machine learning skills would be preferred.
This project is focused on identifying and validating protein biomarkers in blood from patients with various diseases such as postoperative delirium, endometriosis, inflammatory bowel disease, or hepatocellular cancer. We. are using the most comprehensive proteomics platform, the aptamer-based SomaScan, that allows us to measure expression levels of 7,000 proteins in any given sample to identify the discriminating protein biomarkers and fully automated multiplex immunoassays to further validate the biomarkers in larger sample numbers of independent patient cohorts. The goals are to define protein signatures, develop predictor models, generate new insights into the pathophysiological pathways, and identify novel therapeutic targets. The main anticipated outcomes are to predict risk of development of delirium in surgical patients or of endometriosis in premenopausal women, to detect cancers at early stages such as hepatocellular cancer prior to current diagnosis, and to predict and monitor therapeutic response in patients with inflammatory bowel disease. The students will be involved in all aspects of the project and participate in measuring protein expression as well as analysis of proteomic data for biomarker discovery and predictor model development. Students should have primary skills in lab work such as pipetting. Knowledge in statistical analysis or machine learning would be ideal, but not a necessary skill.