We develop computational methods to functionally annotate the human genome. Through integration of high-throughput transciptomic and epigenomic data from a variety of tissue and cell types, we predict context-specific functional non-coding DNA elements.
An unsupervised learning framework to predict functional DNA elements in the human genome.
Predict tissue-specific functional regions in the human genome through integration of data from Roadmap Epigenomics Project.
GWAS and sequencing downstream analysis
We develop statistical frameworks to better analyze genetic data and dissect the genetic architecture of complex traits and diseases. In particular, we focus on applications of integrative genome annotations in downstream analyses of GWAS and sequencing-based association studies.
Genetic risk prediction
We also develop novel statistical methods to improve the accuracy of genetic risk prediction. In particular, we achieve better prediction accuracy through explicitly modeling external information such as linkage disequilibrium, functional annotations, and pleiotropy.