Changed ccp4 files - mystery

[RachelAlcraft]

A slight mystery - the ccp4 files that I have downloaded previously have different values in to the ones I have downloaded today, for example: 6eex used to have 01_NX=53 02_NY=36 03_NZ=87 04_MODE=2 05_NXSTART=-1 06_NYSTART=-9 07_NZSTART=-14 08_MX=24 09_MY=36 10_MZ=64 11_CELLA_X=9.21 12_CELLA_Y=11.98 13_CELLA_Z=22.8 14_CELLB_X=90 15_CELLB_Y=90 16_CELLB_Z=90 17_MAPC=1 18_MAPR=0 19_MAPS=2 20_DMIN=-1.2534446 21_DMAX=6.23184 22_DMEAN=0.003896585 ISPG=19 NYSYMBT=320 EXTTYP=0 NVERSION=0 ORIGIN_X=0 ORIGIN_Y=0 ORIGIN_Z=0 MAP=MAP RMS=0.8009706

But now has 01_NX=36 02_NY=24 03_NZ=64 04_MODE=2 05_NXSTART=0 06_NYSTART=0 07_NZSTART=0 08_MX=24 09_MY=36 10_MZ=64 11_CELLA_X=9.21 12_CELLA_Y=11.98 13_CELLA_Z=22.8 14_CELLB_X=90 15_CELLB_Y=90 16_CELLB_Z=90 17_MAPC=1 18_MAPR=0 19_MAPS=2 20_DMIN=-1.2536675 21_DMAX=6.2328725 22_DMEAN=2.1361067E-10 ISPG=19 NYSYMBT=320 EXTTYP=0 NVERSION=0 ORIGIN_X=0 ORIGIN_Y=0 ORIGIN_Z=0 MAP=MAP RMS=0.79792327

I do not yet know what impact this has.

[RachelAlcraft]

I had a very helpful reply from Chimera support:

Comment:

If I download PDB 6q53 and the electron density server map from PDBe for 6q53 in Chimera 1.16 I get the atomic model is positioned outside the map.

 open 6q53
 open edsID:6q53

The atomic model is located in an adjacent unit cell of the X-ray map. You can use symmetry to get the part of the adjacent unit cell map that would cover the atomic model with Chimera command

 vop cover #0 atom #1

Or you could use the Chimera Unit Cell tool (menu Tools / Higher-Order Structure / Unit Cell to make symmetry related copies of the atomic model that lie inside the EDS map.

So technically the map and atomic model are correctly positioned. But of course it would be more convenient if the EDS map was the unit cell that contains the PDB atomic model.

Based on the CCP4 file header information you report the old 6q53 EDS map had grid size 51 by 57 by 62 while the unit cell has grid size 144 by 144 by 144. The new EDS map has grid size 144 by 144 by 144 and is an entire unit cell. So it appears that the old map only covered and asymmetric unit, or part of a unit cell, while the new map covers an entire unit cell. In some cases having just the asymmetric unit is best and in other cases having an entire unit cell is best, for instance to see what the crystal contacts look like.

It seems the EDS map calculation changed at least for 6q53 if the CCP4 files you have both came from EDS. But both are correct.

In some other software that is specifically for X-ray maps, such as Coot, I think it will automatically apply symmetry to make the map cover the atomic model. Chimera does not do that automatically. So for Chimera use it would be convenient if the EDS map was for the unit cell containing the atomic model.

The behavior with 6q53 is the same in our newer program ChimeraX.

I don't think we can do anything to fix this in ChimeraX (we no longer develop Chimera, it has been replaced by ChimeraX). Because ChimeraX is not specifically an X-ray crystallography program it handles cryoEM and many other maps, and it can have many different atomic models open at the same time. So making ChimeraX automatically apply map symmetry to cover an atomic model has many difficulties because it doesn't know which atomic models belong to which maps. I could perhaps add a command that translates the unit cell map to cover the atomic model. That command would be just as complicated for a user as the "vop cover" command requiring you specify the map and atomic model, but instead of creating a new map would simply shift the original unit cell map. It would not automatically save over the old map, that would require using the existing save command.

Perhaps this explanation will at least help you and PDBe understand the issue. If PDBe offered multiple maps for download, for instance, one that covered just the asymmetric unit of the atomic model and one that gave the unit cell

[RachelAlcraft]

I originally saw this with the small structure 6eex and that went from more than the unit cell to only the unit cell.

I also see a problem if density not feitting the structure in non xray structures, eg the electron microscopy structure 6axz with the density EDS-7017. Coot automatically loads symmetry operations of the electron density into a sphere that covers the structure, but Chimera it does not cover and the comands above for the xray structure do not seem to work.

[RachelAlcraft]

The ultra high res 5d8v is a good example of this. Image the 4 symmetric operations of the strucure - none of them in the density. The density I downloaded a couple of years ago covered the structure.