pafonine at lbl.gov
Tue Mar 31 18:25:11 PDT 2015
MEM = maximum entropy method. This was a quick attempt to implement
MEM-based map modification, as described here (admittedly quite sketchy):
All in all, this works great with very high resolution data (better than
1A), compare usual 2mFo-DFc (blue) vs MEM (red) maps (this is a 0.6A
resolution data set):
To make it work similarly well at lower resolutions more work is
required; so in this sense this is an unfinished project.
FEM = Feature Enhanced Map:
Acta Cryst. (2015). D71, 646-666.
FEM: feature-enhanced map
P. V. Afonine, N. W. Moriarty, M. Mustyakimov, O. V. Sobolev, T. C.
Terwilliger, D. Turk, A. Urzhumtsev and P. D. Adams
On 3/31/15 5:19 PM, mohamed noor wrote:
> Dear developers
> What is the difference between MEM and FEM maps? I read the
> documentation and my understanding is both are supposed to create
> beautiful maps :)
> My problem is I have a structure at 1.9 A (a different protein than
> the one referred to in my previous question), the Rfactor is around
> 20/22 %, Rms bonds is 0.03 and Rms angles is 1.845. If the
> bonds/angles get lower, the Rfactor jumps up (same problem as the
> other structure).
> Looking in Coot, there are positive and negative density and missing
> density for the side chains of residues with high B factor. For some
> reason, phenix.refine keeps on adding water at strange places even at
> regions that I suspect to be just noise. For this reason, I want to
> see if the noise is really noise. A second reason is I have density
> that seem to be too big for a water molecule but waters were placed
> anyway, so I want to know if these are real positive density.
> The dataset was obtained from only 50 degrees, SG is P 61 2 2.
> Including more images brought up the Rmerge to 30-40% and Rfactor
> being stuck at 30 %. For some reason, I don't find the X-ray
> statistics from the phenix.model_vs_data logfile but in essence, I have:
> 28-1.9: 82 %
> 6 A - infinity: 95 %
> in the shells (2.25-2.14, 2.14-2.04, 2.04-1.97, 1.97-1.9) my
> completeness is around 85-62 % and CCwork is at least 70 % and CC1/2
> of the dataset is 60 % at the highest resolution shell.
> phenixbb mailing list
> phenixbb at phenix-online.org
> Unsubscribe: phenixbb-leave at phenix-online.org
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the phenixbb