Joe Cockburn wrote:
Dear BB,
We are trying to solve the structure of a complex between two proteins. We
have two crystal forms of the complex, and a partial MR solution in each.
We now want to average the density between these two forms to improve the
maps, using DMMULTI. However, there are a couple of things I don't
understand.
1. You input F, SIGF, the initial best phase estimate and the Figure of
Merit. But how does DM calculate a map from this? Where does it get the
Fcalc from?
If all you have is the 'best' phase and figure of merit, then there is a
corresponding best estimate for Fcalc, given by
(F_calc, phi_calc) = (fom*F_obs, phi_best)
You can prove that this F_calc will produce the least noisy maps.
2. The program outputs the modified phases and their weights - but again,
how do you calculate a map from this?
Exactly the same way.
For example, in Coot, use Open MTZ (not auto), check the 'Use FOM' box
and select the FOM. So you need to use the columns for FOBS, PHIDM and
FOMDM.
Kevin
