Research Statement Summary
With the advent of high throughput technologies in biomedical research, vast amount of high-dimensional biological datasets have been generated to characterize biological systems and diseases. Particularly next generation sequencing (NGS) technology has being employed by experimentalists to measure genetic, epigenetic and structural changes in DNA and RNA, and generated terabytes of data such as DNA methylation, copy number alteration, mRNA expression and microRNA expression. The availability of these high throughput and complex datasets require new computational methods to be developed to analyze and integrate these data to answer high impact biological questions.
My research goal is to develop open-source computational tools that integrate high throughput biological datasets i) to reverse engineer disease-specific gene regulatory networks and ii) to compute predictive models for biological processes and clinical outcomes.
In my group’s most recent work, we developed a tool to reverse engineer cancer-specific gene regulatory networks in high accuracy. We also developed a Bioconductor package to analyze, cluster and visualize RNA-seq data. I also collaborated in a project where we analyzed RNA-seq data from prostate cancer cell lines to study the role of MST1 localization in prostate cancer development. In the future, I am planning to extend my current work to other types of high dimensional biological datasets such as long non-coding RNA and microRNA expression and build genome-wide gene regulatory networks.
Physical Mapping (PhD Work)
During my PhD studies at the University of California, Riverside, I work in in an NSF-sponsored project whose final aim is to facilitate access to the genes of barley by building genetic and physical maps.
In the context of the barley project, we developed a new method to construct physical maps which exploits genetic markers, and thereby assembles more accurate maps than the ones generated by conventional methods.
We also developed a novel algorithm to compute the minimal tiling path (MTP) of a physical map. Our MTP tool computes the minimal tiling path by using only the restriction fingerprint data and substantially improves the coverage when compared to the MTP produced by currently available tools.