You may also have to change your grid for the fft ("resampling" in photoshop parlance).
From the sfall documentation:
Different spacegroups have special requirements
for factorisation. No prime factors higher than 19 are permitted (The
`atom map' generated on this grid has the asymmetric unit defined for
this spacegroup. See table below .).
The following general restrictions must be observed:
<nx> >= 2 * HMAX + 1
<ny> >= 2 * KMAX + 1
<nz> >= 2 * LMAX + 1
In addition there are further space group dependent conditions as
follows (where `n' is an integer, i.e. <nz> must be even, for
instance):
<nx> <ny> <nz>
P1 - 2n 2n
P21 2n 4n 2n
P21212 4n 4n 4n
P21212a '' '' ''
P212121 '' '' ''
P4122/P4322 4n 4n 8n
P41212/P43212 '' '' ''
P31/P32 6n 6n 6n
P3/R3 '' '' ''
P3121/p3221 '' '' ''
P61/P65 6n 6n 12n
Viswanathan Chandrasekaran wrote:
Hi Fei:
I once had the same error message when using an EM map as an MR model. To make this work,
you need to change your map's axis order from XYZ to whatever is appropriate (in my case
it was ZXY) and also change the spacegroup to P1 using the CCP4 program
"maputils".
In the CCP4i GUI, you do as follows:
1) Go to maputils: "Map and Mask Utilities" -> "Edit/Rotate Maps & Masks".
2) Choose "Edit a map/mask file" for a "map" file.
3) Choose your map as input and choose an appropriate output file name.
4) Under the "Edit file" section, enter "Space group" as 1 and "Change axis order
to" ZXY. You can get this from mapman's header when you load the map or CCP4 log file header from the
failed SFall run. Alternatively, you could try all possibilities in the drop-down list.
SFall should run fine using the output map from maputils.
Best,
Vish
-----Original Message-----
From: CCP4 bulletin board [mailto:[email protected]] On Behalf Of Fei Li
Sent: Sunday, November 10, 2013 4:33 PM
To: [email protected]
Subject: [ccp4bb] help on preparing EM map for mask generation and molecular
replacement
Dear experts,
I'm planning to use a EM map to generate a mask for density modification and
also potentially use it in molecular replacement. But I'm stuck on how to make
this EM map to behave now.
The original EM map was a reconstruction from 2D helical tube. The file was
huge (containing several layer of the protein) and ended with .map. Our
computation collaborator had already cut out a dimer, which is what I'm
interested in, and put it into a small unit cell just slightly bigger than the
dimer density. this file is ended with .mrc.
I can open both file with coot and chimera fine. I tried to convert the dimer map into a .mtz file but
when I try to use any of the ccp4 program (sfall, mapman etc), I always get error message on something
like "map and mask grid samplings do not match", "Fatal disagreement between input info
and map header" or "Incorrect sampling grid factors
40 29 45". I'm thinking it may be the unit cell that it is already in. But
as I'm not familiar with Chimera, I couldn't figure out how to remove it. Also, how
to choose and/or specify a proper grid? Any suggestion on how to make this EM map to
work is high appreciated.
Thank you very much!
Best regards,
Fei
Fei LI
Graduate Assistant
310 Biochemistry Building
Department of Biochemistry and Molecular Biology Michigan State University East
Lansing, MI
48824
