On Wed, Jul 21, 2010 at 11:46 AM, Benjamin SCHWARZ schwarz.ben@gmail.comwrote:
> > Does the resolution never appear it map files ? > > Resolution is no stored in the density files. > > What is the incidence/relevance of "arbitrarily" setting this parameter > such as it is apparently done in this example ? > > Setting a wrong resolution will lead to wrong fitting results. In the next > steps of the procedure we simulate the template protein to the resolution of > the map and apply a cross-correlation measure to compare between the two. > If the smoothed template significantly differs from the density map we will > not be able to get correct solutions. > > > If I get it right; though this parameter is critical for computations, the > resolution is not stored in density files. HENCE you have to remember the > resolution of your map, so that you can fill this parameter in the map > header when you load it in IMP. > > In fact, I think I don't really understand the notion of resolution in Cryo > EM, or maybe it is just in IMP. I have an intuition from my experience with > RX structures, where resolution has something to do with the precision and > size of data used for the density map generation; and ultimately represents > some kind of a "level of detail". > > In order to better understand the two notions, I used > IMP.em.SampledDensityMap and IMP.em.write_map to generate density maps of > a structures with different values for voxel size and resolution. It appears > that increasing the resolution of a map (while keeping the size of a voxel > unchanged) indeed blurs the information and "potatoes" isodensities, which > is fine; but for some reason it also affect the data size : the bigger the > resolution, the bigger the map size (in angstrom as well as in number of > voxels). Is this normal a behavior ? >
just to clarify: pixelsize and resolution of an EM map are not necessary correlated - but for computational efficiency one should choose the pixel (or voxel) size according to the resolution. the max resolution of an EM map is 1/2 voxelsize. however, most people will oversample to avoid signal loss as result of interpolations. so voxelsize should be 3-4 * resolution. if the voxelsize of your map is ridiculously small compared to the resolution you can resample your map in a dedicated image processing software, typically using trilinear interpolation (spider, eman, tom-toolbox).
cheers
frido
> --Ben > > _______________________________________________ > IMP-users mailing list > IMP-users@salilab.org > https://salilab.org/mailman/listinfo/imp-users > >