dd@beta.lanl.gov (Dan Davison) (07/06/89)
[the following is commentary on the message that appeared on the
BIONET login banner July 1]
Part of the message said:
As many of you know from our bulletin board postings over the last
year, the BIONET grant was up for renewal. A special study section was
formed to review the Resource and a site visit was conducted.
Unfortunately the BIONET Resource was not recommended for further
funding primarily due to the opinion of the site committee that the
requirement for a strong research program had not been met. Research,
in addition to BIONET's service, training, and communications
functions, is required by the NIH Division of Research Resource
regulations.
If BIONET did not fulfill its obligations under the current contract,
then there can be no quarrel with the termination of that contract.
However, that does not mean that the baby should be thrown out with
the bath water.
After receiving this decision, we explored other methods of obtaining
funding in an attempt to preserve what we had every reason to believe
was a valuable service to the research community. Currently, for
example, BIONET is used for over 60 searches of the various sequence
databases every day of the month!
I infer from this that the study section for Bionet felt that the
service side of BIONET-- the access to databases, the searching
capabilities, and the communications with other scientists-- were
relatively unimportant.
This view is a mystery to me. I don't know who was on the study
section, but I would suspect that they were all computer "experts*".
They would consider the capabilities of BIONET easily reproducible in
their laboratories. But for the great unwashed masses of bench
scientists, this is not true. Yes, restriction enzyme searches should
be done with local computing power. Even with fast database search
algorithms, not every laboratory can afford a fast 386-based PC with
lots of hard disk for the data bases, a phone line or other connection
to the campus network (& therefore the Internet or Bitnet), and the
time to update data regularly. Also, laboratories that I have known don't
have a computer dedicated to databases and searching--usually it has
to be the grant manager, memo writer, and manuscript machine also. It
simply wouldn't be available to do long searches of the databanks.
I think that the NIH has an obligation to the 2000+ users of
Bionet to make an equivalent facility available. Certainly NIH has
raised the expectations of the molecular biologist as far as computers
in everyday laboratory work are concerned.
What would this replacement machine have? I think that it should
(1) contain current versions of all the molecular biology databases,
and allow users to retrieve parts of those databases
(2) serve the communications functions that BIONET now serves,
(3) contain facilities which the average user doesn't have useful
access to presently. [Restriction enzyme searches, and protein
translation can be done on local PCs]. Database searches, protein
and nucleic acid structure prediction, phylogenetic analyses, and
similarly computationally intensive procedures *do* belong on such
a machine.
The only way such a facility would be created is if the NIH
understands that there is a demand for it. The only way they will
know that is if we do something about it. If you are interested in
expressing your desire for such a facility, or have comments about
what programs should be available at the facility, let me know or post
your comments.
BTW, I'm *not* interested in running such a facility, but I do think
we can influence NIH enough so that they would issue a RFP on this
subject.
The alternative is that NIH will pay for hundreds of PCs and workstations
instead of a central facility, not the best use of limited research funds.
dan davison/theoretical biology/t-10 ms k710/los alamos national laboratory
los alamos, nm 875545/dd@lanl.gov (arpa)/dd@lanl.uucp(new)/..cmcl2!lanl!dd
davison@UHOU (bitnet)
As always, I speak for myself, and only for myself.
(*) For reference, I would probably be considered an "expert" also. I
don't need the analysis or database functions of Bionet; but I do
think the communications functions are important. I also know a lot
of scientists who do need the analysis and database functions!
--
dan davison/theoretical biology/t-10 ms k710/los alamos national laboratory
los alamos, nm 875545/dd@lanl.gov (arpa)/dd@lanl.uucp(new)/..cmcl2!lanl!dd