Re: [ccp4bb] PHIDM AND FOMDM from DM

2009-02-09 Thread Eleanor Dodson 

Xie Jiabao wrote:

Dear all,

I am using the density modification tool in ccp4 to generate improved phases 
for/from my model. I find that the electron density map I generate using Fobs, 
and density modified phases (PHIDM) are not the same as that generated using 
Fobs, phicalc (original calculated phases) and FOMDM (new improved figure of 
merit post dm). Can someone please explain to me why this is so?

Thanks in advance,
Xie



  
  

The calculated map is a fourier transform of h k l W*F, PHI

So if youwant a DM style map you need to input
F=Fobs W=WDM PHI=PHIDM . Density modifiction will change phases and 
weights..
( I think DM may also put out a term FDM which equals WDM*FOBS so you 
could use F=FDM, PHI=PHIDM))


I dont know what PHICALC is but if you use that phase you must use its 
associated weight

Eleanor

Re: [ccp4bb] PHIDM AND FOMDM from DM

2009-02-06 Thread Clemens Vonrhein 
Dear Xie,

you need to use the right columns for calculating maps. From your
description it seems you could do the following in FFT:

  a) map after density modification:
  
   F1=Fobs PHI=PHIDM W=FOMDM

 or (DM should write out these columns as well):

   F1=FDM PHI=PHIDM

 See: http://www.ccp4.ac.uk/dist/html/dm.html#labout
  
  b) map before density modification (I assume you have some
 f.o.m. attached to your phicalc)

   F1=Fobs PHI=phicalc W=FOMcalc

Doing

  F1=Fobs PHI=phicalc W=FOMDM

combines the wrong bits of information - and probably just shows you
the map before density modification (since the phases haven't
changed).

A map in the end is calculated with an amplitude and a phase. The
f.o.m. is a way to generate that amplitude by doing (Fobs*fom) - which
should give you the same thing (the amplitude of the centroid).

Usually all density modification program (or refinement programs for
that matter) will write out specific map coefficients to calculate
maps with. So in the same way REFMAC will give you FWT/PHWT for a map
you get FDM/PHIDM from DM.

If you always use these map coefficients (usually described in the
documentation) you'll be doing the right thing and won't get confused
with mixing up different amplitudes and phases (adding f.o.m. for good
measure).

I can happily live without the need to use a FOM-column in any step
during structure solution or refinement, which avoids a very common
mistake of using a weight on an already weighted amplitude (F1=FDM
PHI=PHIDM W=FOMDM ... brrr!). The columns in an output MTZ file are
nearly always very well documented in the program writeup.

Hope that simple explanation helps ...

Cheers

Clemens


On Thu, Feb 05, 2009 at 09:43:02PM -0800, Xie Jiabao wrote:
 Dear all,
 
 I am using the density modification tool in ccp4 to generate improved phases 
 for/from my model. I find that the electron density map I generate using 
 Fobs, and density modified phases (PHIDM) are not the same as that generated 
 using Fobs, phicalc (original calculated phases) and FOMDM (new improved 
 figure of merit post dm). Can someone please explain to me why this is so?
 
 Thanks in advance,
 Xie
 
 
 
   
-- 

***
* 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)
***

[ccp4bb] PHIDM AND FOMDM from DM

2009-02-05 Thread Xie Jiabao 
Dear all,

I am using the density modification tool in ccp4 to generate improved phases 
for/from my model. I find that the electron density map I generate using Fobs, 
and density modified phases (PHIDM) are not the same as that generated using 
Fobs, phicalc (original calculated phases) and FOMDM (new improved figure of 
merit post dm). Can someone please explain to me why this is so?

Thanks in advance,
Xie