THE MAIZEGDB GENOME BROWSER (Guidance group | Beta testers )
Based upon the 2006 MaizeGDB Working Group Report (available at the bottom of this page) and the Allerton Report, it has become evident that the focus of MaizeGDB must be shifted to better accommodate a sequence-centric paradigm. In order to:
1) show how the data at MaizeGDB relate to the maize genome
The feasibility of implementing of a genome browser at MaizeGDB was investigated and various available software solutions were evaluated.
2) relate MaizeGDB's current sequence data with other sequence information as it becomes available
3) become the keeper of maize's "official" set of gene models (which ensures proper nomenclature)
4) create a way to compare the various assemblies and annotations simultaneously
Each genome browser software has such strengths and weaknesses, so determining which software best suits the needs of maize geneticists is a task that requires careful analysis.
Because the maize community communicates well, has a clear vision of their research problems, and has good ideas on how best visualize a sequenced maize genome (and because we believe that the next generation genome browsers should not be specified by a secluded group of programmers), we prepared a survey to gauge cooperators' impressions of existing software and to find out what sorts of functionalities they would like to have in a maize genome browser.
PREPARATION OF THE SURVEY
The MaizeGDB team worked with the MaizeGDB Working Group and the Maize Genetics Executive Committee to prepare and disseminate a survey.
The Number of Respondents
From the 1,241 cooperators contacted, 99 responded.
Time Spent Accessing Maize Online
37% percent of the survey takers spend an hour or two each week online to access maize data online. 39% percent spend between two and five hours. 15% spend more than 5 hours online to access maize data. Only 8% of the survey takers did not use online maize data resources.
Genome Browsers Used
66% of the respondents use MaizeSequence.org and the number is the same for Gramene users. A total of 75% use MaizeSequence.org and/or Gramene (which both use the Ensembl software). TAIR (GBrowse), MAGI (GBrowse), and PlantGDB (xGDB) are being used by 54% of the respondents (but not always by the same people). 42% use the NCBI Map Viewer (software not available for use outside NCBI).
The features are sorted as follows (rankings are shown in parentheses where a lower number indicates a higher priority for the user): Ease of use (1.9), visuals (2.6), speed (3.2), cross-species comparison (3.7), multiple gene selection (4.1), differentiation between computational and experimental data (4.1), and ontologies (5.1). Clearly, the respondents want an intuitive genome browser that allows researchers locate the needed data in the most accessible and fastest fashion.
The most desired feature section of the survey is very helpful to guide genome browser developers in the creation of new features. The users want to reach specific data in the most intuitive tools. They also want downloadable data sets in various formats. The respondents are in need of enhanced cross-referencing between different websites. They desire the most current data and the tools that are easy to learn and to apply for their specific research needs. In short, the users want the minimized hassle and effort in reaching the needed maize data.
"Bad Genome Browser" Examples
Among 29 comments left in "Bad genome browser examples" cite either MaizeSequence.org or Gramene (66%), which use Ensembl as a genome browser. The reason might be that Maizesequence.org or Gramene is the most used browser for the maize cooperators (75% of the respondents uses either site), but this does not seem to be the case given that PlantGDB, TAIR, and MAGI are also highly utilized (each is listed by 54% of the respondents as being used) but PlantGDB is mentioned only once as a bad browser, MAGI only three times, and no one cited TAIR or as hosting a bad browser. Note that both TAIR and MAGI use GBrowse software. For MaizeSequence.org and Gramene (Ensemble-based browsers), the respondents usually cite perceived slowness of the website as the major (and sometimes the only) problem.
Based upon results of the Genome Browser Survey, we support the use of GBrowse for the following reasons:
1) In the "Feature ranking", the three most desired features listed as: ease of use, visuals, and speed. Cross-species comparison capabilities (where Ensembl shines) is only ranked 4th, and GBrowse now has such capabilities available (Synbrowse, CMap, etc.).
2) As indicated in the "Indispensible features" cooperators would like to see specific tool development in the genome browser to enhance their research (e.g., finding genes between two markers). Therefore, a genome browser chosen by MaizeGDB should allow high flexibility in terms of code and tools development as well as community involvement. The flexibility of tool development is intrinsic feature of GBrowse, which allows customizable plug-in architecture, because it defines itself as a community-based open source project.
3) Sites using GBrowse had significantly fewer negative comments than sites using other mainstream software platforms.
4) MaizeSequence.org/Gramene is already providing maize sequence information using Ensembl. Providing these data via GBrowse would allow researchers to use different genome browsers for different applications. For example, when a cross-species comparison across many clades is necessary, Gramene would be most likely be accessed. When it comes to displaying data to answer maize-specific research problems, GBrowse likely would stand out given that customized code could be developed for MaizeGDB and subsequently contributed to the open-source GBrowse project. Offering these two browsers to use by maize researchers will facilitate answering different research problems and will enhance research overall.
5) Interlinks can be provided between different genome browsers, so choosing browser software different from what MaizeSequence.org and Gramene use will enrich the research toolbox at a minimal cost.
ROADMAP FOR IMPLEMENTING THE BROWSER
We are currently working to implement GBrowse at MaizeGDB.
We plan to choose 5 people to provide guidance (we will make sure that they are from academia, industry, and outside of the US), and 10 people for beta testing among the cooperators who agreed at the end of the survey to be a part of the Genome Browser implementation.
We plan to write proposals to include new analysis of the maize genome in structural genomics and systems biology that includes protein structure models and pathway representations. Note that for pathways, collaboration with Gramene would be reasonable given their current focus on pathway data and tools.
For more detailed information on the survey, survey results, or the path for implementation of GBrowse at MaizeGDB, please contact us using the feedback button at the bottom of this page.
GUIDANCE GROUP (in alphabetical order)
- Peter Balint-Kurti
- Sarah Hake
- Damon Lisch
- Mike Muszynski
- Virginia Walbot
BETA TESTERS (in alphabetical order)
- Alain Charcosset
- Olivier Dugas
- James Estill
- David Hessel
- Damon Lisch
- Mike Muszynski
- Paul Scott
- Virginia Walbot
- Rachel Wang
- Cesar Alvarez-Mejia
Return to the homepage
Last updated 10:03 am, Aug 23, 2010.