A.F.W.Coulson%EDINBURGH.AC.UK@CUNYVM.CUNY.EDU (10/03/88)
I'm not sure that this was an invitation to respond, but assuming
it was:
I don't think there is much within the area of molecular-biology-
narrowly-defined which requires this sort of support. There is plenty of
good software available for multi-caacccecess systems for sequence-related
computing which is not very expensive, and DES, via SERC, is already
spending significant sums to make a complete range of these programs
available for use at Daresbury at no cost to UK academic and related
research workers. There is _some_ highly priced software in this area,
but it's not worth subsidising.
The position is slightly different for single-user machines (ie, in
practice, PC's, AT's and clones) because the best available software is
expensive even to academic users. Would support for this type of program
stumble over the necessity that it should be available on a range of
machines? Another potentially serious problem is that there is a number
of competing packages, and opinions differ very sharply about which the 'best'
is. The parallel with UNIRAS may break down at this point, because I believe
that there was reasonably widespread agreement amongst computer services that
they all needed something like UNIRAS to provide a particular type of service
to their clients, and that UNIRAS would be suitable if available cheaply
enough. For molecular biology software, it is the end-users who are going
to decide what software they want, and I should be very surprised indeed if
there was a widespread consensus about which packages are even adequate,
and even more surprised if this consensus survived for more than 6 months.
May I suggest that there is a related area in which the kind of support
which has been given to UNIRAS could be very useful indeed? This is the area
of molecular modelling and molecular graphics. The best software here has been
developed in a commercial environment to meet the needs, principally, of
medicinal chemists in commercial companies,a and it is very expensive for
academics to contemplate. Software which is available through academic
exchange -- FRODO, GROMOS for examples of each area -- is intended for
use by specialists; mere molecular biologists tend to find it unfriendly
and/or unsuitable, but they can't afford to buy the commercial software which
is more oriented to their needs and skills
Andrew Coulson