Our lab does research in two complementary activities. First, we develop computational methods to analyze high-throughput genomic data, such as new tools and pipelines. Second, we develop software to better interact with--analyze, visualize, reproduce, scale, and share--high-throughput genomics data. Below are active and past projects in the lab.
With the widespread adoption of high-throughput DNA sequencing technologies in biomedical research, there is an increasing need for computational platforms that simplify and provide broad access to multi-step, compute-intensive analyses of genomic data. To address this need, the Galaxy project has developed a Web-based platform for accessible, reproducible, and collaborative analysis of high-throughput genomics data, aptly named Galaxy (try Galaxy now using our public server). Our lab participates in the Galaxy project, with a focus on Galaxy’s collaboration, visualization, and visual analysis features. A primary goal of this work is enabling visualization of very large genomic datasets on the Web. We are also developing visual analysis applications that combine visualization with analysis and workflows so that visual inspection can be used to guide and workflow usage.
Web-based Interactive Visual Analysis
Our lab develops frameworks and applications for doing interactive visual analysis on the Web. Visual analysis combines visualization with analysis tools & pipelines so that visual inspection can be used to guide tool & pipeline usage. One aspect of this work is enabling visualization of very large genomic datasets on the Web, and another aspect is integrating visualizations, tools, and pipelines in a meaningful way.
Translational Cancer Genomics
Because cancer is a disease of the genome, genomic analyses have proven useful for understanding the disease state and suggest potential treatments. Galaxy pipelines for tumor analysis combine multiple genomic features (e.g. mutations, structural variations and rearrangements, gene expression levels) into a comprehensive genomic profile of a tumor. Next, the pipelines integrate publicly available data with private patient data to find promising drugs and similar laboratory models that can used for high-throughput screening. By virtue of integrating these pipelines into Galaxy, they provide standardized, reproducible, and understandable pipelines for clinical cancer genomic analyses.
Venomics and Parasite Infection Strategies
This work uses a combination of transcriptomic (RNA-seq) and proteomic (Mass spec) methods to identify the proteins in parasitic wasps’ venom. This is joint work with Nate Mortimer, and manuscripts from this work appear in PLoS One and the PNAS.