Hu JINAddress: Department of Biomedical Informatics, Harvard Medical School, 10 Shattuck St, 3rd Floor, Boston, MA 02115, United States.
Mobile: (+1)217-898-7356 E-mail: Send me an e-mail by clicking the button at the top right corner of the page. Find me at Google Scholar. |
I am a postdoctoral fellow at Dr. Peter J Park's lab at Department of Biomedical Informatics, Harvard Medical School. I am working on the development and application of computational methods for integration of three-dimensional chromosomal interaction data with cancer genomics data.
EducationPh.D. in Physics, Aug. 2013 - May 2018
Bachelor of Science in Physics, Sept. 2009 - Jul. 2013
Publications
News
Archive |
Research highlights
Unified framework for nucleosome positioning
Nucleosomes form the fundamental building blocks of eukaryotic chromatin. The precise role of DNA sequence in governing the genome-wide distribution of nucleosomes has been controversial. We have developed a unified computational framework for modeling genome-wide nucleosome landscape, based on the statistical mechanics model of one-dimensional hard rods and the cross-entropy method for optimization. The proposed framework is able to account for various factors that affect nucleosome deposition, including different DNA sequence features (10-bp periodicity, G+C content, poly(dA:dT) tracts, etc.), steric exclusion between neighboring nucleosomes, and active remodeling by chromatin remodelers. Using this approach, we have characterized the contribution of distinct sequence features to different aspects of the nucleosome landscape, settling several long-standing debates in the chromatin field. |
RNA-binding proteins in embryonic stem cells
While years of investigation have elucidated many aspects of embryonic stem cell (ESC) regulations, the contributions of post-transcriptional and translational mechanisms to the pluripotency network remain largely unexplored. In particular, little is known in ESCs about the function of RNA-binding proteins (RBPs), the protein agents of post-transcriptional regulation. Our collaborators at the Robert Blelloch lab performed an unbiased RNAi screen of RBPs in an ESC differentiation assay and identified two related genes, NF45 and NF90/NF110, whose knock-down promoted differentiation to an epiblast-like state. With a leading role in bioinformatics analysis, I developed a method based on Poisson processes and sequential importance sampling for identifying binding targets of RBPs from RIP-seq data, as well as a gene set expression analysis method for dissecting functional interactions of several genes within a small regulatory network. Our work elucidated the role of NF45 and NF90/NF110 in regulating ESC pluripotency and differentiation. |
Periodicity in nucleosomal DNA
Certain periodic nucleotides in the DNA sequence have been suggested to facilitate nucleosome positioning since the 1980s. However, the extent of this periodicity in nucleosomal DNA and its precise role in directing nucleosome positioning still remain unclear. We carried out rigorous spectral analysis on recent high-resolution nucleosome mapping data and provided a clear and detailed answer to the long-standing question in the chromatin field over the past 35 years about periodicity in nucleosomal DNA. |
Paused pluripotent state in mouse ESC
Embryonic diapause is a reversible state of suspended development triggered by unfavorable conditions. Our collaborators at the Miguel Ramalho-Santos lab found that partial inhibition of mTOR, a major nutrient sensor and promoter of growth, induces reversible pausing of mouse blastocyst development ex vivo. As the leading bioinformatician of the project, I applied a location-scale method for batch effect removal, as well as global correlation analysis in gene, pathway, and gene ontology levels, to compare RNA-seq data generated in three different labs. Our bioinformatics analysis showed that the paused ES cells induced by mTOR inhibition mimic natural diapause transcriptionally. |