| Quick navigation: | Home | Site Map || References | Biography || Copyright | Other copyright | Contact us | Advert | | |
Re: [ccp4bb] combining AutoMR and brute |
||
- Protein crystallographyMain steps:- Protein purification- Crystallisation Special:- Programs for crystallography- X-ray detectors Basic tutorials:- Chemistry- Protein - Peptide - Amino Acids Xtal community:- CCP4BB |
CCP4bb navigationCCP4bb <-- 1999 <-- November 1999 <-- 30 November 1999Subject: Re: combining AutoMR and brute From: Randy Read rjr27 {- at -} CAM {- dot -} AC {- dot -} UK Date: 2009-12-22 No, we haven't made that possible, largely because we can't see circumstances where this would be desirable. However, you can get as close as you want to a brute force search by changing the criteria for rescoring the results of the fast search. If you include the command FINAL TRA SELECT PERCENT 65 (changing from the default of PERCENT 75), then all fast translation peaks greater than 0.9*65=58.5% of the maximum difference from the mean will be rescored, and, after rescoring, everything above 65% of the maximum will be kept. You can do this from ccp4i on the line labelled "Final selection criterion for TRANSLATION search peaks". If you only want to adjust which fast translation peaks are rescored but not which are kept (which is closer to running the brute translation search), then you can include the command FINAL TRA STEP 1 SELECT PERCENT 50 to rescore everything above 50% of the maximum from the fast translation search. To do this from ccp4i, you have to open the "Additional parameters" pane, then under "Criteria for the translation search" change "Rescore criterion". Instead of PERCENT, you can also specify the criteria for which peaks to keep by SIGMA (i.e. Z-score), NUMBER or even ALL. So if you said FINAL TRA STEP 1 SELECT ALL this would essentially be like using the brute translation search, except that the sampling of points is one a fractional grid instead of the more efficient hexagonal-close-packed grid used for the brute translation search. My feeling is that this would just be a waste of computer time. I don't think we've seen a case where rescoring anything below about 50% of the maximum has made a difference -- but would appreciate hearing otherwise if someone has had a different experience! Best wishes, Randy Read On 22 Dec 2009, at 00:25, Francois Berenger wrote: > Hello, > > Is there any way to search in AutoMR mode but have the > translation only being brute forced? > > Regards, > F. ------ Randy J. Read Department of Haematology, University of Cambridge Cambridge Institute for Medical Research Tel: + 44 1223 336500 Wellcome Trust/MRC Building Fax: + 44 1223 336827 Hills Road E-mail: rjr27@cam.ac.uk Cambridge CB2 0XY, U.K. www-structmed.cimr.cam.ac.uk CCP4bb navigationCCP4bb <-- 1999 <-- November 1999 <-- 30 November 1999 |
|
| ProteinCrystallography.org: Copyright 2006-2010 by Quid United Ltd |