[Chimera-users] accuracy of fitting atomic structure in density map at different resolution

[Tom Goddard]

Hi Andy,

  Do you mean how much fits vary depending on EM map resolution?  I suppose you could apply Gaussian filtering to the EM map to produce a lower resolution EM map, then fit your atomic model, then see how much motion and rotation there is of a fit at one resolution from a fit at another resolution.  All that is easy in Chimera.  To compare different fits I would open multiple copies of the same atomic model and fit one to each different map then use the "measure rotation" command (e.g. "measure rotation #2 #3") to see how much motion there is between fits.

	http://www.cgl.ucsf.edu/chimera/docs/UsersGuide/midas/measure.html#rotation

With any other fitting software you could do the same test by exporting the Gaussian smoothed maps from Chimera, fitting with other software, writing out fit PDBs with the other software and then loading those PDBs in Chimera and comparing their orientation.  In that case you'd want to use the Chimera "match" command to measure the motion between models (e.g. match #2 #3 showMatrix true, the look in Reply Log for results).  The "measure rotation" only reports coordinate frame motions done within Chimera, whereas "match" aligns two atomic models.

	http://www.cgl.ucsf.edu/chimera/docs/UsersGuide/midas/match.html

  Tom


On Sep 11, 2013, at 7:13 AM, Anindito Sen <emailanindito at yahoo.co.in> wrote:

> Hi Tom,
> 
> Is there any method by which, we can figure out the accuracy of docking of an atomic structure into its EM density map generated at different resolutions and by different software packages? 
> 
> Thanks
> 
> Andy
>  
>> Dr. Anindito Sen (Ph.D)
>> Department of Cell Biology & Anatomy
>> Graduate School of Medicine
>> University of Tokyo
>> Tel & fax: +81-3-5841-3339
> 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://plato.cgl.ucsf.edu/pipermail/chimera-users/attachments/20130911/13de0c92/attachment.html>