[sharp-discuss] How to interpret each of the data columns in the final output MTZ file of the autoSHARP?

[Clemens Vonrhein]

Dear Jiawei,

to calculate an electron density map, you need an amplitude and a
phase. You only need a weight (the f.o.m.), if the amplitude is some
kind of 'native' amplitude.

All results from SHARP (or the following density modification step)
will use alreade pre-weighted amplitudes. So the only reason why we
keep columns of f.o.m. (FOMsha, FOM, FOMshasol etc.) in the final MTZ
files is for historical reasons - mainly. But there also might be
external programs you want to use after SHARP/autoSHARP that can't
deal with the normal way of encoding phase probability distributions
(HL coefficients).

So whenever you want to calculate a map, you only need to assign two
columns in FFT:

  LABIN F1=Fsomething PHI=PHIsomething

To summarize, what is in eden_flat_xx.xpc.mtz (see also
http://www.globalphasing.com/sharp/manual/appendix1.html, which needs
a bit of updating):

  Fcentsha, PHIcentsha       - centroid map (light atoms only) from
                               SHARP 

  Fcentshasol, PHIcentshasol - centroid map (light atoms only) from
                               SOLOMON

  FBshasol, PHIBshasol       - centroid map (including average HA
                               contribution) from SOLOMON

  Fdm, PHIdm                 - centroid map from DM step (after
  SOLOMON)

Some more remarks:

 - there are more columns in the file, but these can either be ignored
   or are only there due to the extended description we're using. See

     Bricogne, G., C. Vonrhein, C. Flensburg, M. Schiltz and
     W. Paciorek. Generation, representation and flow of phase
     information in structure determination: recent developments in
     and around SHARP 2.0. Acta Crystallogr D Biol Crystallogr, 59(Pt
     11):2023-30, 2003.

 - you have to decide, what kind of map you want to look at: should it
   contain the (average) contribution of heavy atoms or not? Imagine a
   Br-SAD experiment: do you want a map with the Br atoms or without
   (i.e. 'light atoms only'?).

   If the model you are going to build into the electron density map
   will be refined against a Br-containing dataset, you want to use
   the FBshasol/PHIBshasol columns (so that the map you are buuilding
   in shows the Br atoms). If you have a dataset without the Br atoms,
   then use the Fcentshasol/PHIcentshasol columns.

   The HL coefficients always contain the average HA contribution:

     HLA-HLD are from SHARP

     HLAdm-HLDdm are from the DM step after SOLOMON (this is the only
     reason why we do a DM step: since SOLOMON/SIGMAA don't write out
     the HLs of the density-modified map we have to run a DM step at
     the end. Because you might want to feed-back the density modified
     phases into another round of HA-parameter refinement).

 - The FPsha/SIGFPsha columns are amplitude and sigma of the so-called
   'unperturbed structure factor': these will always contain the
   average heavy-atom contribution.

 - the easiest is to use the tools we provide through the interface
   for viewing maps: you can even just save the map(s) as a tar
   file. This way you can see what columns are used for which map.

As a simple guideline:

  SAD :  F1=FBshasol PHI=PHIBshasol
  MAD :  F1=FBshasol PHI=PHIBshasol
  MIR :  F1=Fcentshasol PHI=PHIcentshasol
  SIR :  F1=Fcentshasol PHI=PHIcentshasol


All this is really simple, once you get used to all the relations
between HL, FOM, F, PHI and the light-atoms structure and the HA
substructure.

If you (or anyone) wants to know what columns to use from which MTZ
file for a specific task - please let me know and I'll try to add more
explanations to the manual.

Hope that helps ...


Cheers



On Sun, Jul 18, 2004 at 05:12:21PM +0800, Jiawei Wang wrote:
> Hi, everyone:
> 
> There are the following columns in the final mtz file
> "eden_flat_xx.xpc.mtz":
> 
> Fcentsha, PHIcentsha, FOMsha, HLA, HLB, HLC, HLD
> 
> FHsha, PHIHsha
> 
> FreeR_flag
> 
> FPsha, SIGFPsha
> 
> Fcentshasol, PHIcentshasol
> 
> FBshasol, PHIBshasol
> 
> PHIshasol, FOMshasol
> 
> FCshasol, PHICshasol
> 
> PHIdm, FOMdm, Fdm, HLAdm, HLBdm, HLCdm, HLDdm
> 
> 
> I have several questions:
> 
> 1. The electron density map after SHARP is calculate by FFT using the
> column: FPsha, SIGFPsha, PHIcentsha, FOMsha. What does the column
> Fcentsha mean? How to use it?
> 2. The electron dentsity map after solomon is calculate by FFT using the
> column: FBshasol, PHIBshasol. Is it possible to calculate the ed map
> using the columns FPsha, SIGFPsha, PHIshasol, FOMshasol?  Are these two
> methods equivalent?
> 3. Do the dm columns PHIdm, FOMdm, Fdm come from the dm results of SHARP
> map, or SOLOMON map?
> 4. " LABIN FDM=FBshasol PHIDM=PHIBsha" is displayed in the "final DM log
> file", but I can't find any column labeled with PHIBsha in the
> eden_flat_xx.xpc.mtz file.
> 
> Thank you very much.
> -- 
> Jiawei Wang <jwwang at cryst.iphy.ac.cn>
> Institute of physics, Chinese Academy of Sciences
> 
> 
> _______________________________________________
> sharp-discuss mailing list
> sharp-discuss at globalphasing.com
> http://www.globalphasing.com/mailman/listinfo/sharp-discuss
> 

-- 

***************************************************************
* Clemens Vonrhein, Ph.D.     vonrhein AT GlobalPhasing DOT com
*
*  Global Phasing Ltd.
*  Sheraton House, Castle Park 
*  Cambridge CB3 0AX, UK
*--------------------------------------------------------------
* BUSTER Development Group      (http://www.globalphasing.com)
***************************************************************