Please excuse my previous empty message which was not intended as a
droll comment:
clemens suter-crazzolara <suter2 at urz.unibas.ch> writes:
> ... if i have a protein, and i need to get the sequence,
>i tend to leave the sequencing to a specialist, rather then to get
>involved in the chemistry myself. this is true for a lot of different
>methods related to biology (overexpression, even production, cell-
>culture, somitimes even actual cloning). we tend to give a lot
>of jobs away. now, if i had to go to courses on all these subjects,
>including computing, and had to get involved in each region of
>interest, there wouldn`t be much time left for experiments.
>>what i want to bring across is the general idea that we shouldn`t
>expect everybody to get to deeply involved in computing (as we
>also shouldn`t when it comes to protein sequencing).
This is one of the core questions: just how much computing does a
molecular biologist need to know? The point is well made that knowing
how to operate a computer is quited different from knowing something
about computation, just as knowing something about protein sequencing is
different from knowing how to operate a protein sequenator. My
perception is that, currently, broad conclusions about the function,
evolution and structure of macromolecules are being made based on
sequence similarity and inferred homology. Few graduate programs
provide the training to critically evaluate such claims, and many people
are left to either take the claim on faith or to reject it as
unimportant. Similar points can be made about many other
computationally intensive aspects of molecular biology such as predicted
structures of RNA, X-ray and NMR structure determinations, molecular
dynamics calculations, phylogenetic analysis, etc.
The point is that people need to be aware of enough of the background of
techniques, such as those listed above, that impinge on their speciality
that they can critically evaluate the published results. I think that
there are very few subspecialties of molecular biology that are not
influenced in a major way by the techniques mentioned above. This
doesn't mean that one needs to be able to run the programs, or even the
computer, but it does mean that you need to know something about the
algorithms and their limitations. My personal feeling is that some
general information about computational approaches and the kinds of
problems that are tractable, as well as those that are intractable,
provides a useful context. It may not however be useful for everyone.
Michael Gribskov
gribskov at ncifcrf.gov
