Life Science Informatics Market Struggles in 2006 as Pharma Tightens the Purse Strings

After a slight reprieve in 2005, the life science informatics sector resumed its long-term consolidation trend this year, as three high-profile players — Lion Bioscience, Tripos, and Inpharmatica — were all sold during 2006.

 

Lion’s bioinformatics business, which had been on the block for nearly a year, was sold to BioWisdom in April for €4 million ($5 million) [BioInform 04-07-06]. Tripos, meanwhile, spent most of 2006 mulling a number of options for its informatics and discovery research groups before announcing in November that it had sold the informatics business to Vector Capital, a San Francisco-based equity fund, for $25.6 million [BioInform 11-27-06]. Then in December, drug-discovery firm Galapagos announced plans to buy Inpharmatica in an agreement valued at a minimum of €12.5 million ($16.6 million) [BioInform 12-08-06].

 

Gene Logic also opted to divest a portion of its business during the year, but not its informatics business. Wrapping up a strategic review that it began in June, the company last week announced that it had sold its preclinical division to San-Francisco-based contract research organization Bridge Pharmaceuticals for $15 million in order to focus its resources on its drug repositioning business and to explore “strategic alternatives” for its genomics database business.

 

The M&A pattern is likely to continue into 2007, at least according to one company. Accelrys CEO Mark Emkjer said several times this year that the firm plans to build on the success of its 2004 acquisition of SciTegic by “aggressively” pursuing business development and M&A activities in the months ahead [BioInform 10-27-06]. In August, Emkjer said that the company had identified 55 potential acquisition candidates, though he did not disclose the names of any of those firms.

 

Meanwhile, informatics firms are selling their wares in an increasingly tough market. Customer R&D budgets are shrinking, and many biopharmaceutical informatics teams still prefer to build their own informatics infrastructures in-house rather than buy them from a vendor.

 

At a recent conference in Philadelphia, an IT executive from Pfizer said that his group still develops around 80 percent of its applications in-house, while an official from Millennium said that around 80 percent of the company’s bioinformatics tools are developed in-house, and about 30 percent to 40 percent of its cheminformatics tools are developed internally [BioInform 09-29-06].

 

That trend hasn’t gone unnoticed by informatics companies. Last month, Gene Logic CEO Mark Gessler said that in the discovery area, where most of the company’s products and services have been focused, “there has been a dramatic decline in the investment by pharmaceutical companies over the last five and six years.” [BioInform 11-10-06]

 

The deal flow with large pharmaceutical firms certainly appears to be slowing down. Life science informatics firms disclosed 16 licensing deals with top-20 pharmaceutical firms in 2006, compared to 21 deals disclosed in both 2004 and 2005 (see table, below). Ingenuity and GeneGo accounted for six of those 16 agreements, indicating that demand for pathway informatics is still relatively healthy relative to the rest of the sector.

 

With the discovery informatics market in the doldrums, many firms are trying to move their products further downstream. Accelrys’ Emjker said that the firm’s acquisition strategy would likely be targeted at companies that could help it expand into drug development and clinical research, while Gessler said that Gene Logic is “evaluating strategic options for our genomics capabilities and assets in areas such as clinical biomarker development and molecular diagnostics.”

 

Geospiza also sees an opportunity for its software in the clinical setting. The company recently raised $3 million in financing that it plans to invest in marketing its Finch data-management software to the genetic testing sector [BioInform 12-08-06].

 

Other firms are concentrating on biomarker discovery, buoyed by the FDA’s release in March of its Critical Path Opportunities List, which detailed a number of high-priority projects under the Critical Path Initiative, including biomarker identification, which was at the top of the list [BioInform 03-24-06].

 

In June, statistical software firm SAS added genomic and proteomic analysis capabilities to its SAS Drug Development software platform in order to allow customers to integrate biomarker and molecular information with their clinical analysis methods. At the time, Laurie Rose, director of global health and life science for SAS, said that the Critical Path Opportunities List “was a perfect fit for that direction that we had already decided we wanted to go in.”

 

Other firms, like Insightful, Rosetta, and Genedata, took steps during the year to adapt their statistical gene expression analysis methods to gene and protein biomarker analysis.

 

Even Microsoft threw its hat in the biomarker ring this year. In April, the company formed the BioIT Alliance, a network of life science hardware and software providers created to improve biomedical data exchange using Microsoft technologies. In October, the initiative launched the so-called Biomarkers Project to address interoperability challenges in the field of biomarker identification and validation [BioInform 10-20-06].

 

Partners in the Biomarker Project include Affymetrix, Accelrys, Agilent Technologies, Amylin Pharmaceuticals, Applied Biosystems, the BioTeam, Digipede Technologies, Discovery Biosciences, Geospiza, Hewlett-Packard, Illumina, InterKnowlogy, Sun Microsystems, Strand Life Sciences, Transenda, and VizX Labs.

 

 

While many familiar players in the informatics sector are eyeing downstream markets for their products, some new faces emerged during 2006 to address traditional discovery informatics challenges. In particular, the field of bioinformatics witnessed a renaissance of sorts for accelerated hardware to speed genomic database searching.  

 

More than a decade after companies like TimeLogic, Paracel, and Compugen pioneered the use of reconfigurable hardware in order to accelerate compute-intensive bioinformatics algorithms, a new crop of players has sprung up to take advantage of recent developments in field-programmable gate array programming as well as alternative hardware-based acceleration methods.

 

Companies like CLC Bio, Mitrionics, Progeniq, and Adaptive Genomics have all released FPGA-based products over the last year targeted at the bioinformatics market. In November, Mitrionics kicked off the Mitrion-C Open Bio Project, an initiative to create an FPGA development “ecosystem,” according to Anders Dellson, CEO of Mitrionics. [BioInform 11-10-06]  

 

Dellson told BioInform at the time that the bioinformatics sector is particularly attractive for FPGA vendors because its strong open source roots “make it one of the first verticals that can quickly benefit from FPGAs.”

 

Some vendors are also eyeing the bioinformatics market as an early adopter for accelerated applications based on graphical processing units, which were developed primarily for the gaming market. This fall, Stanford University’s [email protected] project released a version of its client software that runs on GPUs from ATI Technologies. In a presentation to announce the launch, Stanford’s Vijay Pande said that the software running on ATI’s Radeon X1900 graphics card “can achieve almost 100 gigaflops per processor.” [BioInform 10-06-06]

 

Another alternative accelerator is the Cell Broadband Engine processor, developed by Sony, Toshiba, and IBM for use in Sony’s PlayStation 3 and IBM’s BladeCenter QS20 server. The processor combines a general-purpose processor with multiple GPU-like coprocessors.

 

IT shop Mercury Computer Systems is partnering with researchers at Boston University to accelerate computational drug-design algorithms using the Cell processor. So far, they have ported an application called fragment-based drug design, or FBDD, to the processor and have seen a 10-fold speedup compared to BU’s Blue Gene supercomputer in a chip-to-chip comparison [BioInform 11-27-06].

 

Unlike most IT firms looking to test their technology in the bioinformatics market, Mercury eschewed sequence-alignment algorithms for its proof-of-principle for the Cell. Mirza Cifric, director of Mercury’s biotech group, said that Blast, Smith-Waterman, and the like are already “well-suited for general-purpose processors” and wouldn’t see the “tremendous advantage” that FBDD demonstrated when it was ported to the Cell.

 

Of course, that doesn’t mean that there’s no need to accelerate sequence searching. As of Dec. 21, GenBank contained nearly 67 billion base pairs — a 19 percent growth rate over last December (see table, below). As biological databases continue to grow, there will always be a need for bioinformatics developers to devise new ways to efficiently sort through that data.