[Chimera-users] Creating maps from models for map/model FSC calculations
ardan at ebi.ac.uk
Fri May 27 01:20:49 PDT 2016
Dear Tom and Elaine
The suggestion to use the vop resample command works very well - many thanks
@Tom: If we could have the script which does the map on the second grid without interpolation that would be very useful. We would like to run FSCs as well between the different ways maps are being derived from models and it would be interesting to see to what extent the interpolation affects things at different spatial frequencies.
Many thanks and best wishes
On 26 May 2016, at 19:26, Tom Goddard <goddard at sonic.net> wrote:
> Hi Ardan,
> Sounds like you want the molmap calculated map to have a grid that exactly matches a second map which is a cube centered on the molecule. The way to do that is compute the molmap in the normal way then use “vop resample” to interpolate the molmap map on exactly the grid of the second map. This is more reliable than trying to figure out the molecule center and make a matching cube map without using the second map. I can think of at least 3 reasonable ways to define the center (equal weighted atoms, mass weighted atoms, bounding box center) and it is just too easy to get a grid that doesn’t really match the second map. So if #0 is your molecule and #1 is your second map use
> molmap #0 5 gridSpacing 1.3 model #2
> vop resample #2 onGrid #1 model #3
> then map #3 will exactly match the grid of map #1. The interpolation is trilinear (using nearest 8 grid points) and will be very close to the exact values you would obtain by using the second map grid to sum the atom Gaussians. If you really want the exact sum of Gaussians on the second map grid with no interpolation step I can give you a Chimera Python script that will do that.
>> On May 26, 2016, at 12:43 AM, Ardan Patwardhan wrote:
>> Dear all
>> We are using Chimera to align PDB models to maps, then creating maps from the models using the following command (as an example):
>> molmap #0 5 gridSpacing 1.3
>> This should create a map with the voxel spacing of 1.3A and a gaussian atom spread function with the width related to 5A resolution (https://www.cgl.ucsf.edu/chimera/docs/UsersGuide/midas/molmap.html).
>> Our problem is
>> a) the map that is created can have an arbitrary size - we need it to be cubic and have the same (known) dimensions as the second map against which it will be compared
>> b) we have tried padding it in other programs like bsoft but then we sometimes run into problems of centring the model.
>> I was wondering if there was a way to get a centred map of a certain size from a model directly from Chimera?
>> Perhaps there are certain parameters that can be used in molmap?
>> You thoughts and advice on this issue are much appreciated!
>> Many thanks and best wishes
>> Ardan Patwardhan
>> Coordinator - cellular structure & PDBe production
>> Protein Data Bank in Europe (PDBe)
>> European Bioinformatics Institute (EMBL-EBI) European Molecular Biology Laboratory
>> Wellcome Trust Genome Campus
>> Hinxton, Cambridge CB10 1SD
>> Tel: +44 1223 492649
>> Chimera-users mailing list: Chimera-users at cgl.ucsf.edu
>> Manage subscription: http://plato.cgl.ucsf.edu/mailman/listinfo/chimera-users
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the Chimera-users