IUBio Biosequences .. Software .. Molbio soft .. Network News .. FTP

Bio-informatics query

Bob Phair rphair at ix.netcom.com
Fri Aug 22 18:48:14 EST 1997


Each sector of the bioinformatics/computational biology community seems
to have its own definition of the field. In my view, there is one
important chasm that must be bridged if we are to make bioinformatics a
success. That bridge must be built between those who consider
bioinformatics a part of computer science and those who consider it a
part of biology. If bioinformatics remains tied to database and pattern
recognition/sequence analysis tools, and fails to encompass mechanistic
modeling, that bridge will be difficult to build. 

Biologists think mechanistically. Medicine, as Lewis Thomas has
eloquently emphasized, is at its best when based on mechanism. If
bioinformatics speaks only the language of pattern recognition and
statistics, bench biologists and physicians will not comprehend our
field. Even quantitative genetics is closer to statistics than it is to
biology. Those of us who watched the last 30 or 40 years of
"interactions" between experimental and theoretical biologists have seen
this pattern played out over and over again. The two disciplines exist
in separate intellectual worlds, effectively recreating CP Snow's Two
Cultures. Effective interaction is extremely rare. There are separate
professional societies, separate (though sometimes parallel)
professional meetings, and separate journals. This is fine for
developing academic careers, but it has subverted any chance that
experimentalists and theoreticians might commonly work together toward
common goals.

Bioinformatics is presently building toward separatism. It has the fresh
air of early successes in its lungs. Bioinformatics is already strong
enough to proceed without the counsel or participation of working
experimental biologists. Bioinformatics is, for example, already
developing wonderful international conferences where computer scientists
talk to computer scientists and biologists talk to biologists. If it's
not broken soon, this habit will set like concrete, and both groups will
be forever excluded from each other's inner circle. In the early days of
theoretical biology, it became fashionable for biologists to dismiss
theoreticians as ignorant of experimental data and difficulties, and
theoreticians to dismiss experimentalists as ignorant of basic
mathematics. Today, we risk creating a similar undercurrent of mistrust.
If we fill the chasm between bioinformatics and experimental biology
with such a current, the bridge will never be built.

The world is a big place and there is plenty of room for all the
approaches humankind can invent, but your committee could make a truly
lasting contribution by beginning to build this bridge.  You ask how
interactions between computer scientists and biology/genetics
researchers can be fostered. I have suggested to the US National
Institute of General Medical Sciences a program that would supplement
existing biomedical research grants to pay part of the salary of an
experienced theoretical biologist or computer scientist to help
formulate and solve the bioinformatics and computational biology
problems that surround many first-rate research projects. I still think
this idea has merit, and I hope you will consider it as well.

Robert D. Phair, Ph.D.  rphair at ix.netcom.com
BioInformatics Services  http://www.bioinformaticsservices.com
Partnering and Outsourcing for Computational Biology

More information about the Comp-bio mailing list

Send comments to us at biosci-help [At] net.bio.net