Integrative analysis of multi-omics data; integrative cross-species analysis of multi-omics data; AI and machine learning method development for and applications to bio-medical fields; image analysis.
My overall goal is to apply and develop computational tools that analyze biological data to better understand the mechanisms of complex diseases, with a particular focus on AI and machine learning methods. As the Associate Director of Machine Learning & Imaging, Computational Sciences, The Jackson Laboratory, I strive for deep understanding of the most recent machine learning theories and development and identifying their applications in biomedicine, as well as developing new methods for improved analysis. I participated in the MODEL-AD project during the past five years which aims to develop and evaluate various mouse models for human late onset Alzheimer’s disease (AD). When I applied deep learning methods to extract MRI image features and linked them to genetics and metabolites for an AD study of the MODEL-AD project, I realized existing tools are insufficient to enable us to know which mouse strains best reflect the AD mechanisms in human. Hence, my group recently developed a graph neural network-based cross-species gene expression data analysis framework that aims to maximize the similarity between mouse disease models and human disease data. The longer-term goal of my group is to develop an expandable AI framework for cross-species analysis of multi-omics data which not only include genomics, transcriptomics, proteomics, metabolomics, epigenomics, and microbiome data, but also images, including cell images, MRI, fMRI, DTI, CT and multiplexed tissue images.
I have published more than 45 peer-reviewed papers in various journals, including first or co-first author publications in Nature Genetics, American Journal of Human Genetics, Machine Learning, Journal of Computer and System Sciences, and IEEE Transactions on Information Theory.