> From: Alex Brown abrown@nimr.mrc.ac.uk > Date: Wed Sep 25, 2002 09:37:02 am Europe/London > To: Bozidar Yerkovich bozidar@salilab.org > Subject: Re: SEQUENCE_SEARCH problems > > Hi. > > I tried your suggestion of running my machine overnight. The > SEQUENCE_SEARCH took > 6 hr, 15min, 1.32sec, almost reducing my PowerBook to a heap of glowing > titanium :). > > The results were slightly different to those published, but I take it > this could be due to database updates, etc. I did, however, get a > series of error reports in the .log file : > > addgrp__223E> Array too small. Increase MAXSEQ > current maximum, current need: 40 46 > addgrp__223E> Array too small. Increase MAXSEQ > current maximum, current need: 40 41 > addgrp__223E> Array too small. Increase MAXSEQ > current maximum, current need: 40 42 > addgrp__223E> Array too small. Increase MAXSEQ > current maximum, current need: 40 42 > addgrp__223E> Array too small. Increase MAXSEQ > current maximum, current need: 40 41 > addgrp__223E> Array too small. Increase MAXSEQ > current maximum, current need: 40 41 > addgrp__223E> Array too small. Increase MAXSEQ > current maximum, current need: 40 45 > > I can't find any refernce to a variable MAXSEQ in the manual. > > Cheers, > > Alex. > > On Monday, September 23, 2002, at 02:43 pm, Bozidar Yerkovich wrote: > >> >> >> Alex, >> >> 1. MODELLER ftp services are down indeed. I apologize for that. We had >> some >> security problems, but they have been solved. Those files should be >> accessible again by the end of a day. >> >> 2. SEQUENCE_SEARCH can indeed take a long time. It depends on your >> machine's >> CPU and RAM. Just let it run, maybe in the evening. If it's still >> running >> next morning that something is wrong. It should finish in 2 hours or >> so, but >> then again this might be way off for your particular computer. >> >> 3. I am sure if you try google, you find them somehow. >> >> God luck, >> Bozidar >> >