Cross-omics meta-profiling

With the advancement of high throughput genomic, proteomic, and metabolomic technologies, a growing number of omics data sets are becoming available for integrative analysis and provide new resources to solve questions about disease. We have established computational methods to meta-analyze genomic (RNA arrays, microRNA arrays, RNA seq, Genome Seq, Exome Seq), genetic, phenotypic, and other sources of molecular and environmental data. An example of this method includes Gene-expression Based Genome Wide Association Study (eGWAS) for drug/biomarker discovery (PNAS 2012, Diabetes 2015, Diabetologia 2015), Envirome-wide Association Studies (EWAS) to identify gene-environment interplays (Human Genetics, 2013), and transcriptome-wide time course analysis (Nature Immunology 2009 & Clinical Immunology 2008).

Therapeutic target for Type 2 Diabetes

Type 2 diabetes (T2D) is a polygenic disease characterized by insulin resistance. Growing evidence has indicated the causative link between adipose tissue inflammation and the development of insulin resistance. We used eGWAS to study T2D. We discovered and then verified our top functional candidate, the immune receptor CD44, which plays a novel causative role in the development of T2D. We demonstrated this role using CD44 -/- mice, human serum measurements and immunohistochemistry, and through a successful preclinical study of a prototype drug against CD44. This work identified CD44 as a possible therapeutic target for T2D and a novel biomarker for insulin resistance (PNAS 2012, Diabetes 2015, Diabetologia 2015). This work was also highlighted in the feature articles of other scientific journals (Nature 2012, Nature Reviews Drug Discovery 2012).

Neoantigen in Type 1 Diabetes

Type 1 diabetes (T1D) is an autoimmune disease that destroys the insulin-producing β-cells in the pancreas. We recently created a T1D cross-omics meta-profile to identify new autoantigens for this disease. We identified genes/proteins that are dysregulated in the thymus and peripheral lymph nodes of T1D patients, and specifically expressed in pancreatic islets. Vitamin D binding protein (VDBP) was one of our top autoantigen candidates. We performed verification experiments using NOD mice and human samples (Diabetes 2016). There was stronger T-cell reactivity to VDBP compared to controls. Serum anti-VDBP autoantibodies occurred at higher levels and higher frequencies in T1D patients than healthy controls. Immunohistochemical localization revealed that VDBP was specifically expressed in α-cells adjacent to β-cells of pancreatic islets. Thus, we identified VDBP as a neoantigen in T1D.

These study projects have been funded by the government agencies such as NIH and NLM as well as non-profit medical research organization, HHMI, LPFCH, and JDRF.