Scientists led by a team from Wellesley College have developed a genome browser called G-nome Surfer that lets multiple users manipulate and organize genomic information with their hands using a touch screen system developed by Microsoft.
The software is implemented on the Microsoft Surface platform — a tabletop computer that responds to touch and other objects — using C# and the Microsoft Surface software development kit.
It lets users compare, annotate, share, and relate heterogeneous genomic data using publicly available scientific information, gene-expression data, and Blast results by simply touching the computer's screen. Users can also home in on portions of genomic maps and compare multiple sequences side-by-side or by overlaying them on top of each other.
Specifically, users can find genes on a chromosome or search for a gene name or GenBank number; access GenBank notes and publications from PubMed; and view alignment and translation in prokaryotic and eukaryotic chromosomes.
G-nome Surfer also includes a local database containing FASTA and GenBank sequence ﬁles and uses the National Center for Biotechnology Information's Blast+ tool suite for sequence comparison and analysis.
The developers presented the latest version of the software, G-nome Surfer 2.0, at the BioVis conference on biological visualization held in Providence RI, in late October.
BioVis was part of a larger week-long visualization conference hosted by the Institute of Electrical and Electronics Engineers.
BioVis' goal was to acquaint software developers with current visualization programs as well as the challenges of crafting tools to adequately capture and display information contained in life science datasets. Simultaneously, the conference organizers aimed to inform biologists about currently available programs like G-nome Surfer as well as efforts to develop software that could be useful for exploring their data.
The developers said they created the software to help researchers interact with a large amount of heterogeneous genomic information in a collaborative setting, as well as to lower the threshold for using advanced bioinformatics tools.
The current release of the software is meant to introduce students to current applications for genomic analysis and is also intended for use in research presentations, Orit Shaer, an assistant professor of computer science at WC and one of the developers, told BioInform.
However G-nome Surfer isn't meant to replace standard genome browsers like the University of California, Santa Cruz, Genome Browser or the European Bioinformatics Institute's Ensembl. Rather, Shaer said, it "complements" these resources because "it supports [research] collaboration and provides better support for novice users," she explained.
An added benefit is that once beginners have used it, they "will be more easily able to make the transition to more expert tools," she said.
Shaer said her team began developing G-nome Surfer in 2009, "motivated by the observation that emerging human-computer interaction techniques in tabletop computers in particular" could lead to "new possibilities for interaction" with the large datasets that are typical of the genomics space.
Prior to developing the first incarnation of the system, Shaer said her team conducted a user study in which they interviewed about 17 industry researchers, university instructors, and graduate and post doctoral students.
The goal, she explained, was to come up with a list of functional requirements that a piece of software would need to in order to enable users to bring together multiple datasets and tools for their data analysis needs.
Those efforts and the results that informed the initial design for G-nome surfer were described in a paper they presented at the 2010 Conference on Human Factors in Computing Systems.
Once it had developed the software, the team conducted another study in which 50 students working in pairs conducted an observational, behavioral, and anatomical investigation of mice with genetic defects in their neurons.
In the study, the students were supposed to identify which of a set of five genes encoded the mutation that led to the phenotype in the mice.
The teams performed the same analysis using G-nome Surfer on the Microsoft Surface platform, G-nome Surfer on a traditional computer system, and a standard set of web-based NCBI tools.
"In the web-based conditions, users spent significantly more time looking for information, whereas in the tabletop, they spent significantly more time discussing and reflecting on the information," Shaer said.
Additionally, users could organize and then manipulate the information displayed on the tabletop "in a way that made sense," allowing them to, for example, draw conclusions about the role of mutations in the genes they were exploring by comparing gene expression data side by side, she said.
"We didn’t see any of that in the other [two] conditions," she said.
In the next version of the tool, Shaer's team plans to add support for more expert users.
"We are looking at a scenario where [users] will be able to work collaboratively on a Microsoft Surface .... [and then] save the information to a personalized web portal that will allow [them] to access [and use] existing tools, [as well as] ... organize information," she explained.
Currently, interested users can install the software on a standard Windows 7 machine if they don’t have a Microsoft Surface computer, or on a laptop that has touch capabilities, Shaer said.
She added that the team is also working on creating a version of G-nome Surfer that will run on Apple's iPad.
Have topics you'd like to see covered in BioInform? Contact the editor at uthomas [at] genomeweb [.] com.