CHICAGO – This week, a bioinformatics research team at the University of Dundee in the UK received word that the Wellcome Trust will continue funding development of Jalview, software for editing, visualizing, and analyzing sequence alignments, for five more years.
The money wasn't quite what Jalview leader Geoffrey Barton had asked for — Wellcome Trust awards often are not, according to Barton — but it will allow the team to keep working on the software for another five years.
"This is not enough to do everything, though, so we will be going back to other funders this year as well," Barton told GenomeWeb. A grant from the UK's Biotechnology and Biological Sciences Research Council recently expired. Barton does expect to apply for a new round of BBSRC funding later in the year.
The University of Dundee released version 2.11 of Jalview at the beginning of July. This introduced support for VCF files, which are used to annotate variants in sequencing data.
The software could already visualize variants from databases maintained by the European Molecular Biology Laboratory's European Bioinformatics Institute (EMBL-EBI). "This [update] allows you to use your own local variant files and there's more sophistication in how you can interrogate variant data and map it onto sequences and onto three-dimensional structures," said Barton, who heads computational biology at the University of Dundee.
With the new funding, Barton's team will continue an important migration of Jalview from a locally installed software package to a JavaScript version that can run on the web and be embedded in third-party websites.
Development of Jalview actually dates to 1995, when Barton headed genome informatics at the Wellcome Trust Centre for Human Genetics in Oxford, UK. He moved over to EMBL-EBI in 1997, where he led development of the European Macromolecular Structure Database, now known as the Protein Data Bank in Europe.
Barton joined the Dundee faculty in 2001, and the Scottish university has been the home base of Jalview ever since.
Barton and his research group originally developed Jalview to seek out messages that signal genomic specificity in families.
"We realized we needed a way to interactively visualize those kinds of things and draw lots of other data together," he said. "We want to look at it in context with information about three-dimensional structure and other annotations that were available from the public databases."
At its core today, Jalview is for editing sequence alignments, though it also has increasingly become a visualization and analysis tool in recent years.
Barton called sequence alignments a "deceptively simple data structure" that are more complex than they might appear. "You have individual sequences, but then you have relationships between those sequences represented, and then you have calculations that you can make across groups of sequences to infer function and specificity," he explained.
Jalview also can be used for collaboration. "You can share your alignments and your analysis, so if you do some complex analysis, you can save that as a project, which you can then give somebody else and they can open it and they will see exactly what you saw," Barton said. "We have tried to maintain that ability with every new feature, that it can be saved or restored with the project."
The current software incorporates the Jmol and Chimera visualizers, both of which are widely used for visualizing 3D molecule structures. "They integrate very closely with Jalview and that gives you a lot of power to go from sequence-level and evolution-level information through to three-dimensional structures and back," Barton said.
Early releases could run on web browsers or be embedded into other web pages. Barton said that more than 200,000 pages had some form of this Java-based software. But security changes introduced in the last four years or so have made it impossible to run Java applications on web pages.
So Barton's team made the decision to translate Jalview into a JavaScript, a distinct programming language. "We could not do this from scratch since Jalview is complex software with over 300,000 lines of computer code that has been developed over 20-plus years," Barton said.
He turned to chemistry professor and software developer Robert Hanson of St. Olaf College in Northfield, Minnesota, who created a method to translate Java into JavaScript. This new iteration of Jalview, now available in alpha form, is called Jalview JS.
Barton called Hanson's method "magic." To this end, Hanson — who also developed Jmol — appeared in a wizard costume at the Jalview booth at the Intelligent Systems for Molecular Biology and European Conference on Computational Biology conference in Basel, Switzerland, last month.
"JalviewJS simplifies the use of Jalview for people, particularly in teaching, since it does not require any software to be installed on the computer, but the original Jalview can handle bigger problems faster," Hanson explained.
Barton estimated that Jalview has 60,000 to 70,000 regular users worldwide, who have found many different applications for the software.
Tasks have included loading sets of unaligned protein sequences obtained by searching the EMBL-EBI Uniprot protein sequence database; calculation of trees to show relationships between aligned sequences; identification of sequences for which there is a corresponding 3D structure, then displaying these structures overlaid in Chimera or Jmol; and cross-referencing of proteins with EMBL-EBI's Ensembl genome database, then returning the full-length gene and related annotated transcripts and aligned protein products, according to Barton.
"It's used extensively in teaching as well as in research because it's quite easy to get up and running and you can do quite sophisticated things in a simple way," Barton added. In addition to medical schools and universities, secondary schools have started to incorporate Jalview into biochemistry, genomics, and biotechnology classes, he said.
Barton's Dundee group has responded to this growing demand by creating a series of short YouTube videos on how to get started with Jalview in research and educational settings alike. "You can dip in and learn how to use it relatively quickly, rather than to read a complicated manual, although there is a complicated manual as well, of course," he said.
Barton added that his research group has been trying to extend its limited resources by partnering with organizations that seek to bring bioinformatics into schools. Jalview is fully open-source, so outside developers contribute as well.
"It's really a research tool and it's driven by its research goals," Barton said of Jalview. "My view is that Jalview enables you to link from the network effect through to protein structure and then onward to protein-drug interactions.
"My interest is to try to bring those different aspects of research together in a way and navigate between human genetic variation, population variation, and evolution by looking at alignments and 3D structure in a way that is currently a little bit difficult to do."