Projects
Biomedical Data Analytics for Precision Medicine
How to intelligently integrate multi-omics profiles with imaging, clinical, epidemiological, genetical and demographic details, in order to improve the traditional practice of medicine?
Are we able to use advanced deep learning techniques or statistical model to utilize healthcare information in clinical decision-making?
There are a lot of challenging in the electronic health records (EHR) data such as big data sharing/storage, quality control, data analysis and integration with observational patient-level genomic and clinical data in multiple centers, etc.
Statistical Learning in Modern Artificial Intelligence
Statistics is one of the key foundations of Artificial Intelligence (AI). In recent years, the AI-related research has evolved into a multiple interdisciplinary field. Our lab aims to provide some cutting-edge statistical ideas and methods for AI-related algorithms.
Machine Learning in Cystic Fibrosis Study
Cystic Fibrosis is caused by a single gene, but the disease is complex, with symptoms including reduced lung function, a high risk of diabetes, and other consequences of the mutation. Even in the same person, the symptoms may vary over time. With an improved understanding of the biology, individualized treatments might be given, depending on symptom history and genetics. Our lab develops machine learning techniques applied to whole genome sequencing data and other symptom phenotypes for disease prediction .
The Human Microbiome: from Discovery Studies to Statistical Predictive Personalized Medicine
The human microbiome represents a vastly complex ecosystem that is closely connected with our physiology and health. Our lab builds microbiome-based predictive models for personalized medicine. We also develop statistical models and novel methodologies to predict host traits based on human microbiota.