Filip Van Petegem <filip.vanpete...@gmail.com> writes: > While trying to perform some docking experiments with crystal structures into > cryoEM maps, I found that some deposited EM maps are not at the right size > relative to crystal structure coordinates (e.g. a ccp4 formatted deposited EM > map looks smaller than a crystal structure that's supposed to represent less > than 10% of the map). The problem is independent of the graphics program used > (e.g. VMD, chimera, pymol all show the EM map as too small). > > Is there any straightforward way to get the map to the right scale, using > information somehow hidden in the map header? Note that I don't know a priori > by how much the EM map would have to be inflated. I'm using the SITUS > package > for docking - converting from ccp4 to situs, spider or any other common map > format all deliver maps with the wrong size.
Yes, sorry for that - in quite a few maps, the grid spacing is not set correctly during conversion, in particular when converting from some EM formats that do not store a grid spacing in their header. We ask the depositor for the grid spacing they think the map should have - if you download the EMDB XML header file, you can find it in the XML element tags "pixelX" "pixelY" "pixelZ" (parts of the "pixelSpacing" element in "map") in Angstrom. Incidentally, if you find that some maps fail to open in CCP4 programs and others (file too short), try to load and save them in Chimera (as MRC2000 maps) - another conversion problem. This is also one way of changing the voxel size. We are considering moving to MRC2000 as primary format as it seems the only format that properly supports an origin definition in the file. Input about this is welcome, especially if you are interested in docking. If you find an EMDB map that does not work or has the wrong spacing, please contact me at <b...@ebi.ac.uk>! We are planning to fix the map problems during the next few months, but as most things are, EMDB is not a generously staffed activity and has a lot of legacy problems to deal with, but it is good to see these maps being used and we are committed to supporting it. Best -Christoph -- | Dr Christoph Best <b...@ebi.ac.uk> http://www.ebi.ac.uk/~best | Project Leader Electron Microscopy Data Bank, PDB Europe | European Bioinformatics Institute, Cambridge, UK +44-1223-492649
