Re: [ccp4bb]: Problem to find a suitable solution by MR method

[Ricardo Aparicio <aparicio@if.sc.usp.br>]

***  For details on how to be removed from this list visit the  ***
***          CCP4 home page http://www.ccp4.ac.uk         ***

Dear Yunqing,

I would like to mention that you also need to check what fraction of your protein the
search model corresponds to. To have 50% overall identity is different from having 50%
identity in let say only 30% of your protein. If your protein has a very elongated shape
you may have an addition problem also. In addition, restrict your resolution, it is worth
to try 10-4 A only (if this was not the case).

After this, you could try to remove the most flexible parts from the search model (e.g.
sorting the model.pdb by B-factors: sort -k11 model.pdb). If you want a poly-ala model,
you can use pdbset to get it (EXCLUDE SIDE keyword).

If you have some weak anomalous signal, it could be useful for phasing, to get phase
estimates which could help in the translation function.

Another idea is to try other programs, e. g.,
- Molrep and Beast from ccp4 (you can run them using ccp4i, like Amore)
- EPMR - http://www.msg.ucsf.edu/local/programs/epmr/epmr.html

Hope it helps you.

Ricardo
IFSC-USP/Brazil



1) you need to check what is the fraction that the search model

Yunqing Liu wrote:
> ***  For details on how to be removed from this list visit the  ***
> ***          CCP4 home page http://www.ccp4.ac.uk         ***
> 
> Dear all:
> Does anyone experience or know some structure determination cases by molecular replacement method that AmoRe program helps to find the correct solution after CNS program fails to find a suitable one?
> I have a starting model which is 50% identity and 70% positive to my protein in the primary sequence. In my protein there is a big flexible domain. I fail to find a suitable solution with CNS program with the whole model or the partial model. So, I turn to AmoRe and try to change the variables in the ROTFUN of AmoRe program like following:
> 
> CELL_MODEL <a> <b> <c>            --- I think these parameters define the outer sphere in the Patterson vectors
> SPHERE <Irmax>                     --- I think these parameters define the inner sphere in the Patterson vectors
> 
> But I still can't find a suitable solution to start my model building.  Here, I am eager for any advice from the structure determination communities who ever knows or experiences such problems.
> Great Thanks!
> 
> 
> 　　　　　　　　Yunqing Liu
> 　　　　　　　　yqliu@sibs.ac.cn
> 　　　　　　　　　　2003-04-28
> 
> 
>