Hi, I heard about this scenario of getting a nonsensical map a couple of times before, however I was never able to reproduce this problem myself. Could you please send me (to my email address, not to the whole bb) the phenix.refine inputs so I can reproduce this problem and fix it? Specifically, I need: model and data files, and the exact command that leads to bad maps. All files will be handled confidentially. I guess the problem is in HL coefficients that are used by phenix.refine to output phase-combined map. model_vs_data does not use HL coefficients and this possibly explains the difference. If it is the case then we would need to figure out what happens to HL coefficients and this would require us to reproduce your Autosol run. Thanks! Pavel. On 2/26/10 12:31 PM, Xuewu Zhang wrote:

Hi Phenix developers/gurus, I got a SAD structure, the experimental map (overall_best_denmod..mtz) from Autosol looks great. I then built the model and used phenix.refine to refine the model using the "exptl_fobs....mtz", which worked well as shown by the lowered Rfree. However, the map from the refinement makes no sense at all, totally featureless. Using the same refined model and mtz, phenix.model_vs_data did make the right map. I have tried both phenix 1.5 and 1.6. Any clue? Thanks, Xuewu Zhang UT Southwestern Medical Center ------------------------------------------------------------------------

_______________________________________________ phenixbb mailing list [email protected] http://phenix-online.org/mailman/listinfo/phenixbb