[phenixbb] MD with phenix.refine?

Pavel Afonine pafonine at lbl.gov
Fri Aug 20 16:03:35 PDT 2010

  Hi Ed,

- technically, yes:

phenix.refine model.pdb data.mtz strategy=individual_sites 
target_weights.wxc_scale=0 main.bulk_sol=false simulated_annealing=true 

In addition, you may want to change the temperature, number of 
macro-cycles, etc....

- practically, no, this will all probably not make much sense since the 
restraints target does not have an attractive term (does not account for 
electrostatics). So your MD run will likely unfold your protein. To some 
extent you may correct for this by using secondary-structure restraints 
or your manually defined restraints, but still it's not the same...

MD in phenix.refine is just a tool for optimization and was never meant 
to be a proper MD tool.


On 8/20/10 3:53 PM, Ed Pozharski wrote:
> Silly question:
> is there some way to run what would essentially be an MD simulation with
> phenix.refine?  I mean, turn off X-ray-term by setting
> tardy.target_weights.fix_wxc=0, and then run cartesian dynamics?
> Ed.

More information about the phenixbb mailing list