[phenixbb] Ordered solvent distances

Pavel Afonine pafonine at lbl.gov
Wed Aug 1 12:50:48 PDT 2012


Hi Abhinav,

> The default minimum distance that a water gets placed near 
> protein/water atoms using 'ordered_solvent=true' option in 
> phenix.refine is 1.8.
> Isn't this too close?

Possibly. Though we pick waters into Fourier maps (which are different 
from exact electron density distribution due to finite number of Fourier 
coefficients = finite resolution) so the peaks may be shifted. The 
smaller threshold distance is meant to account for this. As a side 
effect it increases the risk of interpreting bad pick. How big this risk 
(given all other peak selection criteria) - that I don't know: can take 
a few days to investigate.

Also, it's a parameter - you can change it to match your taste.

> Also, if two waters can't get placed in the density due to distance 
> criterion, can phenix place one water in dual conformations in that 
> density?

Currently not, but would be good to have, I agree.

Pavel



More information about the phenixbb mailing list