|Title||Research Assistant Professor|
|Institution||School of Medicine and Dentistry|
|Department||Biostatistics and Computational Biology|
|Address||University of Rochester Medical Center|
School of Medicine and Dentistry
601 Elmwood Ave, Box 630
Rochester NY 14642
||Integrative cancer biology program of NIH postdoc award | National Institute of Health|
||Excellent research award of Mathematical sciences of Michigan Tech University in 2008 | MichiganTech University|
12/2012 – present: Research Assistant Professor
Biostatistics and Computational Biology Department of University of Rochester: System Biology
As a computational biologist, I am developing computational methods and innovations for cancer and immunology research. Currently, I am collaborating with experimentalist to investigate how the drug impacts cell's molecular pathway. First, with the help of biologist, we chose important proteins which have significant changes under the drug delivery and used Ingenuity Pathway Analysis database to generate the generic signaling pathway. Second, we analyzed the high-throughput proteomics data generated from proteomic data (Reverse Phase Protein Array) and employed the data to infer the signaling transduction pathway described by systems of ordinary differential equations. Third, we employed numerical optimization methods to optimize the key parameters of the pathway by fitting the experimental data.
The inferred biological network is incorporated into my well developed multi-scale and multi-resolution agent based model (ABM) to simulate and predict tumor's progression by C++/Java languages. Multi-scale ABM consists of intracellular, intercellular and cellular levels. The inferred biological network determines cell's phenotype switch in the intercellular scale. The intercellular module of ABM is used to describe cell's behavior and angiogenesis in the extracellular matrix (ECM) with respect to various phenotypes. The chemoattractants and drug, which diffuse in the tissue module of ABM, stimulate and inhibit the intracellular pathway of the cell by binding cell's receptor. Multi-resolution design allows us to classify the millions of simulated tumor cells into active or inactive clusters regarding to biological rules. And then, we can concentrate limited compute resources onto the region of interest of the tumor. In addition, GPU (CUDA) based parallel computing algorithm is employed to increase our compute capacity. Currently, I am employing ABM to simulate the cancer progression under drug delivery, such as pharmacokinetic and pharmacodynamic (PK/PD) modeling.
Also, I am collaborating with Neurology Department of University of Rochester to employ Lasso and survival model to investigate the key genes for Glioma cancer research.
Furthermore, I am developing the agent based model to simulate immune response under the threat influenza A virus (IAV) and use Multivariate adaptive regression spline(MARS) to estimate the key parameters of the agent based model by fitting the experimental data.
8/2008 – 2012/9: Assistant Professor
Computational Science and Engineering Institute and the Department of Mathematical Sciences of Michigan Tech University: Computational Biology
I did mathematical teaching and computational biology of cancer research. Especially, I employed Graphic programming Unit to speed up the classical agent based modeling. As a faculty member, I not only have good communication and collaboration skills, but also rich experience to write NIH, NSF and DoD proposals for my research. One of my proposals was funded by NIH in 2012.
05/2005 – 8/2008: Research Fellow
Harvard-MIT (HST) Athinoula A. Martinos Center for Biomedical Imaging: Complex Biosystems Modeling
People who are also in this person's primary department.