I am an Assistant Professor (tenured track) in the Department of Population and Quantitative Health Sciences at Case Western Reserve University. My research interests center around Bayesian inference and prediction, high dimensional models, and complex structured data, such as brain imaging and metagenomic data. The modeling efforts build up my skills for practicing and improving innovation, conceptualization and generalization. In my PhD thesis, I advanced a Bayesian calibration model to improve the prediction of abdominal aortic aneurysm (AAA) enlargement by using the high dimensional CT image data, resulting in papers published in IEEE Journal of Biomedical and Health Informatics and Journal of Biomedical Engineering. I also gained extensive experience in developing spatial-temporal methods to analyze brain imaging data and to model the progression of Alzheimer's disease, resulting in papers published in Statistical Methods in Medical Research, Health Services and Outcomes Research Methodology, and Electronic Journal of Statistics. In my postdoctoral training, I worked closely with biostatisticians and clinicians on linking quantitative microbiome profiling from different body sites to cancer-related disease outcomes. I gained broad computational expertise in analyzing across data types and studies, and focused on developing quantitative methods and bioinformatic tools to identify biomarkers that can be used in specific immunotherapy or treatment design. I have developed Bayesian zero-constrained regression modeling with structured priors for microbiome feature selection, which was published in Biometrics. I have also developed progressive permutation for multiple testing problems. The paper was published in BMC Bioinformatics. I have had extensive experience in collaborating with medical researchers from different biomedical fields, which resulted in publications and abstracts in well-known journals, such as Cell, Cancer, and Scientific Reports, etc. In this project, I will support the study planning and develop data-driven, intelligent and user-friendly bioinformatics that could robustly identify important signals and generate quantitative insights. In particular, I will build high-dimensional statistical models and variable selection techniques to identify key drivers or biomarkers of oncology diagnosis and treatment. These collaborative works will open opportunities to stimulate my quantitative research in big data modeling, Bayesian method, and computational tools that again promote applications in more broad biomedical science.
We are interested in using high-resolution microscopy and high-throughput genomics to understand genome regulation in human development and diseases.
Research Information
Research Interests
Ninety eight percent of the human genome is noncoding and contains regulatory elements (enhancers, promoters, insulators, etc.) that form a blueprint for human development. This regulatory genome is often a hotspot for genetic alterations (e.g., mutations, indels, amplifications) that lead to human cancer. Recently, DNA regulatory elements and associated oncogenes have been found to co-amplify into extrachromosomal DNA (ecDNA) circles to drive cancer genome evolution and poor clinical outcome. The Xie lab will decipher the structure, dynamics and mechanisms of the cancer regulatory genome and ecDNA by using super-resolution and live cell imaging approaches , aiming to improve the diagnosis, prophylactic and treatment of human cancer.