Navigating protected genomics data with UCSC Genome Browser in a Box.

Maximilian Haeussler,W James Kent,Matthew L Speir,Hiram Clawson,Brian J Raney,Jonathan Casper,Angie S Hinrichs,Ann S Zweig,Donna Karolchik,David Haussler

doi:10.1093/bioinformatics/btu712

Abstract

Summary: Genome Browser in a Box (GBiB) is a small virtual machine version of the popular University of California Santa Cruz (UCSC) Genome Browser that can be run on a researcher's own computer. Once GBiB is installed, a standard web browser is used to access the virtual server and add personal data files from the local hard disk. Annotation data are loaded on demand through the Internet from UCSC or can be downloaded to the local computer for faster access.Availability and implementation: Software downloads and installation instructions are freely available for non-commercial use at https://genome-store.ucsc.edu/. GBiB requires the installation of open-source software VirtualBox, available for all major operating systems, and the UCSC Genome Browser, which is open source and free for non-commercial use. Commercial use of GBiB and the Genome Browser requires a license (http://genome.ucsc.edu/license/).Contact: genome@soe.ucsc.edu

Highlights

Custom data display is a popular feature of modern genome browsers, allowing users to view their own private genome annotations alongside publicly available data
The University of California Santa Cruz (UCSC) Genome Browser (Kent 2002, Karolchik 2014) has supported user-generated custom annotation tracks since 2001, and recently added support for track and assembly data hubs (Raney 2013, Nguyen 2014) that allow the user to host collections of large data files on a local server and configure many details of the visualization. These solutions require the data files to be uploaded onto a web server or stored on a genome browser's servers, a drawback that has become increasingly problematic as data file sizes expand and analysis involves sensitive genomic patient data
The installation package includes the UCSC Genome Browser CGI binaries and a minimal MySQL database needed to store user session information and selected tables. These pre-loaded tables have been selected based on their size and access frequency on the public UCSC Genome Browser website

Summary

INTRODUCTION

Custom data display is a popular feature of modern genome browsers, allowing users to view their own private genome annotations alongside publicly available data. The University of California Santa Cruz (UCSC) Genome Browser (Kent 2002, Karolchik 2014) has supported user-generated custom annotation tracks since 2001, and recently added support for track and assembly data hubs (Raney 2013, Nguyen 2014) that allow the user to host collections of large data files on a local server and configure many details of the visualization. For a website like the UCSC Genome Browser, the large amount of annotation data makes virtualization more difficult – at the time of writing, UCSC’s database includes more than five terabytes of data for the human genome alone This exceeds the capacity of most desktop PCs and is costly to host on cloud-based computation services ($250 U.S./month on Amazon EBS or Microsoft Azure, as of Mar 2014). GBiB, by contrast, provides access to all of the same annotation data while consuming only about twenty gigabytes of local storage

IMPLEMENTATION

CONCLUSIONS