[phenixbb] TLS refinement in phenix

Pavel Afonine PAfonine at lbl.gov
Fri Jul 2 06:19:16 PDT 2010


> TLS is nothing but a cheaper (parameter wise) way to incorporate 
> anisotropic aspects in the refinement when the resolution is not as 
> high as, lets say 1.5A. 

This is one way of looking at this. The other one, that I prefer, is to 
look at this as a way of modeling different kinds of atomic/molecular 
motions, and therefore to have a more sensible physical model for them 
(#30):

http://www.phenix-online.org/presentations/neutron_japan_2009/phenix_japan_part1.pdf

So, ideally, one would use TLS at any resolution.

> Now, it depends on the individual developers how they want to write 
> the output. Refmac does it in the header of the pdb, Phenix does it 
> right after the ATOM record.

I was thinking that the latest versions of Refmac output total ADP too 
(or at least there is an option for this)... Can someone clarify this?

Pavel.





More information about the phenixbb mailing list