corrupt O files...how to rescue???

[Gerard Kleijwegt]

In article <3s617d$4l5 at mserv1.dl.ac.uk>, moumou at titus.u-strasbg.fr (luc  moulinier) writes:
|> >In article <3s4h79$pqr at news.duke.edu>,
|> >bussiere at PROBLEM_WITH_INEWS_GATEWAY_FILE (Dirksen Bussiere) writes:
|> >|> My O database has apparently been corrupted and I've lost all the models
|> >|> on there.  I can't sam_at_out the coordinates onto a PDB file, because
|> >|> it has apparently lost all contact with them.  Does anyone have any
|> >|> ideas??  Thanks!
|> >|>
|> >|> -Dirk Bussiere
|> >|>  'bussiere at abacus.mc.duke.edu'
|> >|>
|> >
|> >first, make sure it's not an NFS problem
|> >
|> >then, try reading it with sam_at_in and tell the program
|> >you're reading a structure in O format
|> >
|> >if that doesn't work, I'm afraid you probably
|> >have lost whatever was in it
|> >
|> >--cd
|> 
|> I just make a suggestion: you may try to write your database as a ascii
|> file, using
|> O>  writ  *   database.tmp  ;
|> and then look in the file if the datablocks beginning by  TOTO_...  (if
|> TOTO is your molecule label)  are there.  Take all of them and read them
|> inside O by S_A_I as an O formatted file.
|> Gerard, am I saying rubbish ?

hmm, no, but it asumes that you're able to read the corrupted
file first - and if that works then there doesn't seem to be
any problem any longer ;-)

--gerard

|> 
|> 
|> --------------------------------------------------------------------------------
|> 
|> Luc Moulinier
|> tel: (+33) 88 65 33 14
|> UPR 9004 de Biologie Structurale du CNRS
|> IGBMC
|> fax: (+33) 88 65 32 01
|> Parc d'Innovation
|> 1 rue Laurent Fries
|> 67404 ILLKIRCH Cedex
|> FRANCE
|> 
|>