Investigators

Summary

The Cancer Genome Project Initiatives have generated a daunting amount of genomic and deep sequencing data for tens of thousands of human tumors. An overarching challenge of this post-genomic era is to identify and recognize the cancer drivers and targets from these big genomic data, especially those that can be therapeutically targeted to improve the clinical outcome. The mission of our lab is to apply a multiple disciplinary approach inclusive of integrative bioinformatics, cancer genetics, molecular cancer biology, and translational studies to identify driving genetic aberrations and appropriate cancer targets on the basis of deep sequencing and genomic profiling datasets. Our research projects are composed of both computational and laboratory components. Our dry lab researches focus on developing innovative and integrative computational technologies to discover causal genetic and epigenetic alternations, viable therapeutic targets, and predictive biomarkers in cancer. In particular, we have innovated a concept signature (ConSig) analysis that employs molecular fingerprints for high-throughput interpretation of the biological function of candidate targets in cancer (Nature biotech 2009). In addition, we have formulated a 'fusion breakpoint principle' that describes the intragenic copy number aberrations characteristic of recurrent gene fusions, thus enabling genome-wide detection of copy number breakpoints generating gene fusions. Based on these principles we further developed a powerful bioinformatics tool called 'Fusion Zoom' that identifies recurrent pathological gene fusions via integrative analyses of RNA sequencing, copy number, and gene concept datasets (Nature Commun 2014). Further, we have discovered the crucial application of ConSig analysis in revealing the primary oncogenes targeted by genomic amplifications, and developed a new integrative genomic analysis called 'ConSig-Amp' to detect viable cancer targets. Moreover we also developed an integrated computational-experimental approach called HEPA-PARSE for the genome-wide detection of clinically important tumor specific antigen (TSA) targets (Cancer Research 2012). Our wet lab researches focus on experimentally characterizing individual genetic and epigenetic aberrations in breast cancer such as recurrent gene fusions, genomic amplifications, and epimutations, as well as qualifying viable cancer targets and predictive biomarkers for the development of precision therapeutics in breast cancer. Our current disease focus is clinically intractable breast cancers, such as luminal B or basal-like tumors. In particular, by applying the FusionZoom analysis to the RNAseq and copy number data from The Cancer Genome Atlas, we have discovered a novel recurrent gene fusion involving the estrogen receptor gene in a subset of breast cancers. This fusion called ESR1-CCDC170 is preferentially present in 6-8% of luminal B tumors -- a more aggressive subtype of estrogen receptor positive breast cancer. To date, this is the first and most frequent gene fusion yet reported in this tumor entity (Nature Commun. 2014). We are now assessing the druggability of this fusion with the goal of developing effective targeted therapy against this genomic target. We expect that our new discoveries will yield novel insights into the recurring genetic abnormalities leading to breast cancer initiation, progression, and therapeutic resistance, and establish viable targets for effective intervention.