Hi Rob. I think I have two out of three offenders - the low identity between the model and the crystallised protein; and the low completeness of the model - I am using just the most conserved domain as a search model. I did noticed that when I repeated the search with a more complete model the expected LLG (at the start of the run) went up. Thanks for clarifying this issue for me. Best wishes, Rafael. ________________________________________ From: [email protected] [[email protected]] on behalf of R.D. Oeffner [[email protected]] Sent: Monday, 27 August 2012 6:54 PM To: PHENIX user mailing list Subject: Re: [phenixbb] Funky message on phaser MR Hi Rafael, Phaser contains new methods to estimate what the expected LLG value will be by the end of the calculation. The expected LLG value depends on 1) the resolution of the data 2) the RMSD of your MR model, 3) the size of the MR model compared to the structure you are trying to solve. For LLG values less than about 60 (I think) there is no certainty that the MR model will be placed correctly and you get this warning. Improving any of the three parameters above will help you avoiding this message. Regards, Rob On Aug 27 2012, Rafael Lemos Miguez Counago wrote:
Dear all,
I am getting this message when running Phaser-MR in Phenix (1.8-1069):
"Warning: Resolution of data does not achieve (half) expected LLG target: placement of one/first ensemble will be very difficult"
The data scales/merges OK. - 99% complete at 2.2A, Rmerge 8.4%. TNCS is detected by Xtriage and Phaser tries to account for that.
Coincidently (?) my molecular replacement solution is not looking very good. Any ideas?
Best wishes,
Rafael.
-- Robert Oeffner, Ph.D. Research Associate, The Read Group Department of Haematology, University of Cambridge Cambridge Institute of Medical Research Wellcome Trust / MRC Building, Hills Road, Cambridge, CB2 0XY www-structmed.cimr.cam.ac.uk, tel:01223763234, mobile:07712 887162 _______________________________________________ phenixbb mailing list [email protected] http://phenix-online.org/mailman/listinfo/phenixbb