CULTURE AND RESEARCH INFRASTRUCTURE 369
10.3.2.2 Publication Venue
Although both biologists and computer scientists in academia seek to make their work known to
their respective peer communities, they do so in different ways. For many areas within computer
science, refereed conferences are the most prestigious places to publish leading research.^73 In biology,
by contrast, conference publications are often considered less prestigious, and biologists tend to prefer
journal publications. Computer scientists often write abstracts in such a way as to entice the reader to
read the full paper, whereas biologists often write abstracts in such a way that the reader need not read
the full paper. In publishing work that refers to a new tool, a computer scientist may be more likely to
reference a Web site where the tool can be found, while a biologist may be more likely to reference a
paper describing the tool.
This difference in publication venues strongly affects attempts at collaboration. Academic biologists
often do not understand refereed conferences, and computer scientists often think of journals as mere
repositories of papers, rather than a means of communicating results. Given that publication is the
primary output of academic research, this disagreement can be very disturbing and can have important
inhibitory effects on collaboration.10.3.2.3 Organization of Human Resources
While it is clear to all parties from the start that that the professional expertise of the biologist is
needed to do good work in computational biology, a view that equates computer science with program-
ming can lead biologists to underestimate the intellectual capabilities on the computer science side
necessary for computation-intensive biology problems. Thus, many biologists who do see rewards in
bridging the interdisciplinary gap (especially in academia) tend to prefer doing so in their own labs, by
hiring postdoctoral fellows from physics or computer science to work on their problems, keeping these
ventures “in the family,” rather than by establishing partnerships with more established computer
scientists.
Such an approach has advantages and disadvantages. An advantage is that postdoctoral fellows
with good quantitative and computational background can be exposed to the art of biological experi-
mentation and interpretation as a part of their postdoctoral training, the result of which can be the
nurturing of young interdisciplinary scientists. One disadvantage is that by engaging individuals at the
beginning of their careers, the biologist is deprived of the intellectual maturity and insight that gener-
ally accompanies more seasoned computer scientists—and such maturity and insight may be most
necessary for making headway on complex problems.
The integration of computational expertise into a biological research enterprise can be undertaken
in different ways. For example, in some instances, a group of computer scientists can work with a group
of biologists, each bringing its own computational approach to the biological problem. In other cases, a
single individual with computational expertise (e.g., a postdoc) can work in an otherwise “pure” bio-
logical group, offering expertise in math, modeling, and programming. A second dimension of integra-
tion is that the bioinformaticist can play a role as a team member who engages equally with everyone
else on the research team or as a “bridge” that serves as intermediate between practitioners of various
disciplines. These two roles are not mutually exclusive, although the first seems to be more common.10.3.2.4 Devaluing the Contributions of the Other
A sine qua non of interdisciplinary work is that intellectual credit be apportioned appropriately.
In some cases known to the committee, the expertise of scientists from a nonbiological discipline has(^73) Such a practice arose because computer science is a fast-moving field, with a tradition of sharing discovery by online
demonstration and discussion. Conferences were originally formed as a way to talk together, in person, and with relatively fast
publication of results for the requirements of academia. In this context, journal publication would have been much too slow.