In article <40sn51$dg4 at lion.embl-heidelberg.de> Thure Etzold <etzold> writes:
>>I see from the System Support manual (pages 43-48) that GCG 8.1 includes
>>programs with names "srssection" "odd" "srscheck" "srsupdate" and
>>"srsbuild" which are exactly the same as the names SRS 4 uses.
>>the version of the 'srs' in gcg is 4.05 ...so it is pretty up to date
Hi Thure. You are going to have to solve the same problem when you install
GCG 8.1. How do you plan to handle the multiple versions of srsbuild
and the programs?
>>What SDL files does it use? For example, does it include prosite ?
>>(the System Support manual example seems to only have Swissprot, PIR,
>>EMBL and GenBank but could just be a truncated list)
>>the gcg version supports only sequence databases ...prosite is not provided
That alone makes it desirable to run with the real SRS indices.
>>Also, the "LookUp" program (actually a front-end to SRS) has a list of
>>libraries in the program manual example. What happens to this at sites
>>with other databases, or without some of the listed databases? (for example,
>>we do not have genbank or "gb_tags")
What does LookUp do that SRS can't? Did they change the output format
How closely coupled are GCG and SRS? Can Lookup simply use standard
SRS indices (with SRS prepared), or does it need source editing and a
rebuild of LookUp?
>i think it should be safe to add more sequence databanks to the gcg
>installation...note that only databanks with gcg sequence format are
>accepted! ...if you have another SRS installation on your site than the
>safest is to keep that site but, of course, the indices from the gcg srs
>installations can be used. The directory with the indices can be
>specified for each databank: attribute "indexDir" of the #libenv record
>in the file srsdb.sdl
So GCG has to run from a separate index directory if a site has any
database defined in SRS as a "Sequence" database but not in GCG format,
NRSUB at the EBI for example (at the Sanger Centre it just happens to have
moved to "Genome" along with other databases). Or have I misunderstood?
Presumably then GCG's LookUp is picking up "/group=@SEQUENCE_LIBS"
*only* and is not checking the sequence format? Or is it able to skip
databases in EMBL flatfile format (for example)?
This seems to imply needing either a completely separate set of indices
for GCG, or a lot of playing around with symbolic links (on Unix).
Peter Rice | Informatics Division
E-mail: pmr at sanger.ac.uk | The Sanger Centre
Tel: (44) 1223 494967 | Hinxton Hall, Hinxton,
Fax: (44) 1223 494919 | Cambs, CB10 1RQ
URL: http://www.sanger.ac.uk/~pmr/ | England