[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: mlphare



***  For details on how to be removed from this list visit the  ***
***    CCP4 home page http://www.dl.ac.uk/CCP/CCP4/main.html    ***



Dear Ingo,

               On Wed, 12 May 1999, ingo korndoerfer wrote:

> 
> why exactly is the refinement of heavy atom sites in mlphare 
> dependent of the order i put my sites in ?

  it should not. It might be helpful if you'd elaborate on your problem.

> i'm not talking about floating point differences, again.
> rather, it seems, that whatever my first input site is,
> does not refine AT ALL. the others look allright.

  Personally I found that mlphare performs only small positional shifts so
that I had the impession that nothing was done. However the log file
contains information about current values, shifts and esd's for each parameter
that is refined so that you could figure out what was done in each cycle.  
If it is really true that your first site was not refined simple typos in
the refinement statement might be responsible. 

>
> - if the sites are no good, why don't they then all blow  
>   up in mlphare ? why always the first one ?
>   dummy sites don't refine, no matter which position they are in,
>   so that is good.
Check the ATREF statement for the first site.
-- Let me get this straight, you said that the first site was not refined
but when you put in a wrong position it did blow up ? If this is the case
it suggests that the parameters shifts are so small that you might have
the impression that it was not refined. How do the b factors behave ? They
usually show larger shifts. 

> - is it stupid to put a single derivative into mlphare ?
>   ( i guess it is, but i whas hoping to get some assessment
>     of the quality of my sites )

Not at all. SIR/AS worked fine in my case. 

> - is this a spcgrp C2 problem ?

I have not had space group C2 but there must be a number of colleagues
out there who can give you advice. I assume that you cannot or should not
refine the y coordinate in C2 because this is a polar space group.

> 
> i am somewhat confident in my sites, since shake and bake,
> as well as rsps (on the sgis that is, not the alphas ... :-)
> find the same set, independently.
> 
  It might be helpful if you posted your mlphare input file on the
bulletin board or alternatively ask a colleague to read it. It might be a
trivial input error. If problems persist there is alway CNS or Sharp :)

Best regards, 

    Lothar


   Lothar Esser

------------------------------------------------------------------
Dr. Lothar Esser 
UT Southwestern Medical Center Dallas
5323 Harry Hines Blvd.
Dallas Texas 75235-9050
E-mail : esser@chop.swmed.edu
------------------------------------------------------------------