Remember you need to look at half the contour height in a mFo map
compared to a 2mFo-DFc map - the same domain should show up but at a
lower relative contour level in both maps.
REFMAC calculates a similar WCNG and you can look at the graph of m
after refinement to see how close it is to 1. If the structure is well
refined except for a missing domain I suspect it will be ~ 1
Eleanor
Hailiang Zhang wrote:
> Actually I cut a small domain from the well-defined structure (just for a
> test). The missing part showed in 2mFo-DFc map but not in both mFo and Fo
> maps, and the mFo and Fo maps are so close so that I wonder whether figure
> of merit generated by SIGMAA helps or not in this situation...
>
> Best Regards, Hailiang
>
>> On Tue, 2010-08-31 at 13:15 -0400, Hailiang Zhang wrote:
>>> Is the difference
>>> between mFo and Fo maps supposed to be very small?
>> For an essentially correct model, yes. The major advantage of (2mFo-DFc)
>> maps is suppression of model bias, so if you don't see much difference
>> then your model is very well refined. For illustration, introduce a
>> systematic error on purpose and see which map gives you better result.
>>
>> --
>> "I'd jump in myself, if I weren't so good at whistling."
>> Julian, King of Lemurs
>>
>>
|