Biomax, Lion Jump on the Text-Mining Bandwagon
Text mining was a hot topic at ISMB this year: A full SIG meeting, two papers, and no less than 22 posters were devoted to the topic. And it’s not just academic groups who are jumping on board. Both Biomax and Lion Biosciences have recently added literature mining to their range of offerings.
Biomax, which had a poster at ISMB describing its text-mining tools, is currently offering the capability as a service offering, sad Victoria Pernide-Lopez, a scientist at Biomax. The company has several customers using the service so far, she said, but the offering has not yet been formally launched.
Lion, meanwhile, announced a collaboration with TEMIS, a text-mining software firm based in Cambridge, Mass., under which TEMIS’ text-mining tools will be integrated with Lion’s SRS and Target Engine technology.
TEMIS will integrate its Insight Discoverer Extractor and Skill Cartridges natural language processing tools into SRS and Target Engine so that scientists can mine the scientific literature or in-house documents for information that can then be analyzed with other biological data in Lion’s applications.
ISCB Seeks Input on Journal Affiliation
The International Society for Computational Biology is seeking input from its membership on whether Bioinformatics, published by Oxford University Press, should remain the official journal of the society.
Phil Bourne, head of the ISCB’s publications committee, announced at ISMB last week that the society “is going through the process of deciding what journal to be associated with,” and has established three main decision-making criteria: the scientific scope of the journal, whether the journal is open access or not, and the role of the society in selecting editors.
Later, he told BioInform that ISCB’s contract with OUP is set to expire soon, which has opened up the opportunity to perform some “due diligence” and explore some other options that may exist.
Bourne said that 10 publishers have so far responded to ISCB’s request for proposals, including a number of publishers who plan to launch computational biology journals in the next year.
“It’s in the society’s best interests to be affiliated with a high-impact journal,” Bourne said. Bioinformatics’ impact factor is currently 6.7, he said, adding that the society is hoping to bring that up to the range of PNAS or Genome Research, which have impact factors in the neighborhood of 10.
ISCB sent out a survey to its members later in the week, seeking input on the criteria that Bourne outlined. Several observers at ISMB noted that the issue of open access is likely to be a controversial topic among the society’s membership.
Bioinformatics currently charges a subscription fee and makes its articles freely available to the public via the OUP website 12 months after publication.
Battle of the Bioinformatics Boxes
Those hitting the vendor exhibits at ISMB 2004 might have noticed a pattern: Apple, IBM, and Sun Microsystems were all offering “bioinformatics-in-a-box” clusters with pre-installed informatics software packages.
Apple’s Workgroup Cluster for Bioinformatics, which became available in January, is a cluster of Xserve G5 servers bundled with the iNquiry software package from the BioTeam that includes more than 200 bioinformatics applications. Liz Kerr, director of the scitech market at Apple, said that there has been strong interest in the product since its launch, particularly from end-user biologists who don’t have the resources to retain a bioinformatics specialist or IT support staff.
Apple also demonstrated X-Grid at the conference, which allows users to harvest and manage compute cycles from distributed desktop machines X-Grid is currently downloadable in beta, and Apple plans to include it as a standard feature in the next version of the OS X operating system, Kerr said.
Sun meanwhile, was showcasing the new version of the Starter Cluster for Bioinformatics that it launched last month [BioInform 07-26-04]. The cluster is based on the company’s V20z Opteron-based server, and comes with Incogen’s VIBE workflow software, Gene-IT’s GenomeQuest sequence search software, iNquiry, and the BioBox suite of public bioinformatics tools co-developed by Sun and the Asia-Pacific Bioinformatics Network.
IBM had its BladeCenter Solution for Bioinformatics on display. The system uses IBM's 1.6 GHz PowerPC 64-bit processors, and comes preconfigured with Blast, Hmmer, Fasta, Blat, and ClustalW. The system can also use 32-bit processors.
Two New UK Bioinformatics Initiatives
Two new UK bioinformatics initiatives made their debut at ISMB 2004: the Biosystems Informatics Institute (BII) and the UK National Text Mining Center, which has identified bioinformatics as its first application area.
BII, based in Newcastle, was actively recruiting at the meeting. The initiative plans to hire 25-30 bioinformaticists to develop software for “post-genomic research,” according to Ian Humphery-Smith, who is the director of the initiative.
Smith said that the effort has £4 million pounds ($7.3 million) in seed funding from the UK’s Department of Trade and Industry and the One NorthEast regional economic development agency.
The goal, he said, is to work with regional companies and universities to develop software that can be commercialized. Humphery-Smith said he’s confident that BII will be able to reach is ultimate goal of spinning out several companies.
The UK National Text Mining Center also launched in June, and it has ambitious goals as well. Funded with £1 million ($1.8 million) from three UK science agencies, the center John MacNaught said that the service was created to develop and support text-mining tools for biomedical research.
The center will be housed in the Manchester Interdisciplinary Biocenter, a £34 million ($62 million) facility that is still under construction, and will be run by a consortium of four UK partner institutions: UMIST, the Victoria University of Manchester, the University of Liverpool, and the University of Salford. This consortium is investing another £800,000 in the center.
The center will also collaborate with the University of California Berkeley, the University of Geneva, the San Diego Supercomputing Center, the University of Tokyo, and the European Bioinformatics Institute.