CULTURE AND RESEARCH INFRASTRUCTURE 343
A challenge for a field as interdisciplinary as this is that incoming students will arrive with possibly
completely non-overlapping backgrounds. Most programs accept a B.S. in computer science, biology, or
math as a prerequisite; to produce a well-rounded computational biologist will require very different
training programs. The University of Colorado’s certificate program in computational biology requires
incoming students to take preparatory classes in “Biology for Computer Scientists,” “Computer Science
for Bioscientists,” or “Mathematics for Bioscientists,” depending on what the student missed earlier in
his or her education.
An advantage of graduate programs is that when communication among faculty of different disci-
plines is good, graduate projects provide an ideal opportunity for students to work in an interdiscipli-
nary environment. In some cases, work with adjunct professors from industry can lead to exciting
projects. On the other hand, if communication between faculty is poor (which may be possible for
reasons described later in this chapter), a graduate student dependent on completing a project (e.g., a
dissertation) can get caught in the middle of a dispute with no way to graduate.
10.2.2.5 Postdoctoral Programs
Postdoctoral programs at the BioComp interface are also varied. Some postgraduate programs are
explicitly aimed at “conversion,” that is, training a fully trained member of one field (usually biology) in
the basic tenets of its complement. For example, the University of Pennsylvania’s postdoctoral program
in computational biology is a master’s degree in computer and information systems, designed for those
with Ph.D.s in biology who need the training. Other programs focus on involving the participant in
research and laboratory work, in preparation for industry or a faculty position, just as in postdoctoral
programs in other fields.
Some programs, such as Duke’s Center for Bioinformatics and Computational Biology, are similar
to graduate programs in that they focus on genome analysis. Others, like the Johns Hopkins’ program in
computational biology, are firmly grounded in genomics but are pointedly reaching out to larger
questions of integrative biology and experimental biology.
In promoting postdoctoral programs at the interface of computing and biology, it will be necessary
to take into account the very different traditions of the two fields. In biology, one or more postdoctoral
fellowships are quite common (indeed, routine) before an individual strikes out on his or her own. By
contrast, the most typical career path for a newly graduated Ph.D. in computer science calls for appoint-
ment to a junior faculty position or a position in industry—postdoctoral fellows in computer science are
relatively rare (though not unheard of).
Two foundation-supported postdoctoral programs have been influential in stimulating interest at
the BioComp interface: the Sloan-Department of Energy (DOE) program and the Burroughs-Welcome
program.
10.2.2.5.1 The Sloan-DOE Postdoctoral Awards for Computational Molecular Biology^9 For 8 years,
the Alfred P. Sloan Foundation and the U.S. Department of Energy (Office of Biological and Environ-
mental Research) have jointly sponsored postdoctoral research awards for scientists interested in com-
putational molecular biology. The purpose of these fellowships has been to catalyze career transitions
into computational molecular biology by those holding doctorates in mathematics, physics, computer
science, chemistry, engineering or other relevant fields who would like to bring their computational
sophistication to bear on the complex problems that increasingly face molecular biology.
Operationally, the program was designed to offer computationally sophisticated young scientists
an intensive postdoctoral opportunity in an appropriate molecular biology laboratory. In most cases,
awardees had strong educational backgrounds in a computationally intensive field, although in a few
(^9) See http://www.sloan.org/programs/scitech_postdoct.shtml.