We develop cutting edge data science tools and methods for applications in precision genomic medicine, and use these methods to impact and improve the way patients are treated in the clinic.
Our Lab's Focus
The focus of our research is to develop computational and statistical tools to investigate core components that contribute to disease prognosis and etiology, and for the accurate determination of optimal diagnostic, prognostic, and therapeutic regimens for individual patients. We are actively developing methods and software tools for data preprocessing, integration, and downstream analysis, and applying these tools in a variety of clinical and biomedical applications. Our work includes a balance between statistical methods development, algorithm optimization, and clinical application. Statistical innovation in our group focuses on the development of clinically motivated tools that integrate linear modeling, Bayesian methods, factor analysis and structural equations models, Hidden Markov models, mixture models, dynamic programming, and high-performance parallel computing. This work has resulted in widely-used tools and algorithms for profiling transcription factors (MAT, MA2C), preprocessing and integrating of genomic data (ComBat, SCAN-UPC), aligning sequencing reads (GNUMAP), developing multi-gene biomarker signatures (ASSIGN), and metagenomic profiling (PathoScope). We have successfully applied our tools in several biomedical and clinical scenarios, ranging from mechanistic studies and to precision genomics.