About: Visualization

Our facility hosts mirrors of the UCSC Genome Browser for visualization. The internal visualization browser resides on a server composed of eight X5355 2.66GHz processor cores, 24GB RAM, and 27.2TB of local storage. The internal browser is directly linked to the LIMS system (as described above) such that sequencing output is displayed, by flowcell, upon completion of initial and any subsequent run analysis. Various levels of downstream informatics analysis and processing that produces data amenable to browser visualization is also displayed, in addition to 3rd party data. A part of the internal browser is set up to provide alternate, password protected views of data to collaborators. Figure 1 shows screenshots from our internal browser for accessing chromatin accessibility tracks from the Roadmap Epigenomics Project. The checkbox grid interface allows for efficient navigation of large collections of closely related datasets; in this case for a variety of fetal tissues from embryos throughout different developmental stages (top), and hematopoietic samples collected on a number of different individuals (bottom).

Figure 1.

Additionally a dedicated high-availability, fail-safe web server cluster hosts a publicly accessible home page for the Roadmap Epigenomics Project. From here one can browse to various pages providing protocols and data standards for the community to download. Links are also provided to Roadmap data portal sites, one of which is a UCSC mirror site that we also host (http://www.epigenomebrowser.org/). The home page is intuitive and meant to cater to the needs of the intended biologists, medical community researchers and the general public. For example, we provide a visual “fetal tissue browser” and a “stem cell browser” that provide clickable access to project data using a easy-to-understand pictoral interface. See Fig. 2.

Figure 2.

The Roadmap UCSC mirror site is supported by two load balancing management servers and two back end servers, each hosting a MySQLdatabase and Apache web server. The two management servers are Dell PowerEdge R610s with Intel Xeon X5667 3.07GHz CPUs with 16 cores, 32GB RAM, and 100GB of local storage. The back end MySQL and Apache web servers are Dell PowerEdge R510s with Intel Xeon X5670 2.93GHz CPUs with 24 cores, 32GB RAM, and 5TB local storage. The Epigenome Roadmap Project relies upon meta-data generated by the Roadmap project's EDACC and SRA/GEO in order to properly display consortia data. The user community can also view ENCODE data simultaneously.