CULTURE AND RESEARCH INFRASTRUCTURE 351
the idea of capitalizing on progress in genetic technologies. Yet because they predate the bioinformatics
boom, they were often late to the game, catching up by heavy investment or by outright purchasing of
other firms that had organically grown the bioinformatics capability. For example, in December of 2001,
Amgen announced that it was buying the bioinformatics-rich biotech company Immunex Corp for $16
billion.^31 Genentech highlights its own bioinformatics capabilities as a key part of the research portfo-
lio.^32 However, while these firms and the pharmaceutical giants are clearly great consumers of
bioinformatics software and human resources, it is less clear to what extent they are performing original
computational biology research.
A second wave of companies was founded in the 1990s, in the era of the Human Genome Project
and the increase in availability of information technology. Millennium Pharmaceuticals, for example,
was founded in 1993 with the goal of being a science- and technology-driven pharmaceutical company,
with a capability for target discovery based on the human genome information being published. How-
ever, most of Millennium’s drugs on the market have come from acquisitions, and the goal of real
rational drug discovery remains challenging. Millennium does have a high-profile leader in charge of
bioinformatics and uses IT for three main functions: bioinformatic inference making, such as identifying
likely functions of novel proteins or the existence of gene expression patterns that correlate with disease
states; chemoinformatics, searchable databases of chemical structure and biological activity; and com-
putational analysis to predict drug candidates’ physiological qualities such as absorption rates, distri-
bution, metabolism, excretion, and toxicity.
Of higher profile is Celera, which Craig Venter founded in 1998 to compete with the publicly
funded Human Genome Project. While genomics experts still argue over his methods, he certainly
found innovative uses for computational and analytic techniques in stitching together the results of his
“shotgun” sequencing method. Regardless of its scientific success, however, Celera has had little com-
mercial success^33 as it turned from sequencing to the potentially more lucrative field of drug discovery.
It still makes money by offering access to its proprietary databases to other biotechnology and pharma-
ceutical companies, but its has given up on its efforts to commercialize its software platform, selling the
Celera Discovery System to sister company Applied Biosystems (both Celera and Applied Biosystems
are owned by Applera Corporation). In addition to the Celera Discovery System, a subscription-based
database, Applied Biosystems offers an array of software for gene sequencing, laboratory information
management, and gene analysis (as well as a variety of instrumentation and reagents).
10.2.4.3 Start-up and Smaller Companies
The area still receives some attention from venture capital firms such as Flagship Ventures, Kleiner
Perkins Caufield Byers, Atlas Ventures, and Alloy Ventures. However, the emphasis seems to be shift-
ing from bioinformatics to a stronger emphasis on biology, including medical devices and drug discov-
ery. Even companies that once positioned themselves as bioinformatics companies now describe them-
selves as being in the drug discovery business,^34 most notably Celera but also many smaller companies.
For companies that concentrate primarily or exclusively on informatics, times are very difficult, in large
part due to the same sort of bubble collapse as mainstream IT faced from 2000 onward.
Analysts blame overinvestment in the area, leading to more companies than the space can support;
companies founded by IT players with insufficient biological knowledge; and increasing competition
from big players such as IBM and HP.
Midsize companies such as Gene Bank and Incyte have a similar business model to Celera, offering
access to proprietary databases, which often contain patented gene sequences. One model that seems to
(^31) See http://www.informationweek.com/story/IWK20011221S0038.
(^32) See http://www.genentech.com/gene/research/biotechnology/bioinformatics.jsp.
(^33) See http://www.fool.com/portfolios/rulebreaker/2002/rulebreaker020423.htm.
(^34) See http://www.bizjournals.com/washington/stories/2002/07/08/newscolumn5.html.