In article <1993Dec16.122143.14629 at reks.uia.ac.be>, przemko at reks.uia.ac.be (Przemko) writes:
> In article <1993Dec15.110256.6885 at comp.bioz.unibas.ch> doelz at urz.unibas.ch writes:
>>the following shall demonstrate to you the use of the Sequence Retrieval
>>System (SRS) software (Thure Etzold, Heidelberg) with the LISTA database
>>Walk around databases, sequences, links, and gather answers to questions
>>which you didn't dare asking before (sorry, that was PR).
> Well, you know the sysmans, the same all over the world (like postal
> workers I guess).
> Could I simply do it over the e-mail?
The Sequence Retrieval System (SRS) is, like most retrieval systems,
a structured tool which could basically be made available as email server
as well. This is the theory. The praxix shows that a query will
immediately come back with a suggestion how to refine the query, and how
to make the answer more detailed and more sophisticated.
As the size of the biological databases increase, the software to access
biological information has to become more sophisticated as well. In
contrast to accesses via a sophisticated, interactive interfaces the
electronic Mail access to a server implies that you send the query,
wait for the result, and receive the answer. Now at the time where
you receive the answer the result might be overdue, or is even no longer
needed because you got the answer elsewhere. Further, how could
a user interface be convenient enough so that there were the option
to parse email input and reoutput a followup?
The future of the database queries must focus on non-electronic mail
access to allow better queries, and more interactive results. Installation
on a local system, as I suggested above, is one of the options. As
SRS works on the GCG formatted databases as well, you do not need to
plan for a specially formatted database of the sequence databases.
All you need is space for the indices. I would be happy to provide detailed
numbers but I guess the forum here is not too much interested in disk space
but rather functionality.
Other options for database retrieval include networks. Well, SRS is
available on a telnet hole, and future might come up with other functionality
which allows for a network access in entirely different fashion.
The main topic of my posting was that the _combination_ of the special-
purpose database LISTA allows questions to ask which have not been
possible before. LISTA is an interesting case as the data arising from
genome projects have rarely fully assigned and featured gene names,
which can be used to target questions for functionality. In supplying the
DNA and peptide homologies, we strengthen the importance of the LISTA
database as with a retrieval tool as suited as SRS you can immediately
conclude whether or not a given sequence will or will not have a
relation to a yeast sequence.
We are open to suggestions, and as I stated in the release notes, the
homologies and SRS files are experimental. If you feel that there
should be features added, or would like to contribute refined data,
or help otherwise to improve the quality of LISTA, please feel free
to issue your suggestions.
Hope this helps,