Actually, zero spacegroup should be tolerated by the CCP4 library now.
We were recently alerted to another problem, caused by non-zero origin in the
EM map header. But I think this is different to the problem of the original
poster.
It's probably fair to say we haven't tested enough against real EM maps, and
perhaps we should.
Perhaps the original poster could make their problem maps available??
Cheers
Martyn
-----Original Message-----
From: CCP4 bulletin board on behalf of Christoph Best
Sent: Tue 7/13/2010 10:20 PM
To: CCP4BB@JISCMAIL.AC.UK
Subject: Re: [ccp4bb] conversion of cyroEM reconstruction from MRC to CCP4
format
Hello,
we have found that there are some minor issues in MRC files using in Em
that can make them unpalatable to the CCP4 map library. You can easily
look at header fields in Python - here is a sample program to quickly
check some of the more common problems.
from struct import *
header=file('your_file_name.map').read(1024)
print 'dimensions=',unpack_from('iii',header,0)
print 'mode=',unpack_from('i',header,3*4)[0],'should be 0, 1, or 2'
print 'space group=',unpack_from('i',header,22*4)[0],'should be 1'
print 'magic=',unpack_from('cccc',header,52*4),'should be (\'M\', \'A\',
\'P\', \' \')'
print 'machine stamp=','%x' % unpack_from('i',header,53*4),' should be 4144
or 4444'
print 'nsymbt=',unpack_from('i',header,23*4)[0],' should be 0'
The issues we typically encountered were bad space groups (zero), wrong
machine stamps (they should be 0x4444 or 0x4144), missing the "magic
bytes" that spell "MAP ", or a value of nsymbt that indicates the
presence of additional symmetry information following the header, which
is then missing.
Chimera ignores most of these issues.
-Christoph
--
| Christoph Best <b...@ebi.ac.uk> http://www.ebi.ac.uk/~best
| European Bioinformatics Institute, Cambridge, UK +44-1223-492649
