Skip to content

Model-based LocScale
Model-based local sharpening

Model-based sharpening is the legacy mode of local sharpening as originally introduced for the first version of LocScale described in this paper. In LocScale 2.0, model-based sharpening can be run in a completely automated procedure, including robust ADP refinement of the input model. Model-based LocScale can still be a very useful sharpening tool in cases where atomic models have been fitted, but the map displays substantial resolution variation.

Model-based LocScale workflow


alt text

Usage

locscale -hm path/to/halfmap1.mrc path/to/halfmap2.mrc -mc path/to/model.pdb -v -o model_based_locscale.mrc

Here, emmap.mrc should be the unsharpened and unfiltered density map. If you wish to use the two half maps instead, use the following command:

locscale -hm path/to/halfmap1.mrc path/to/halfmap2.mrc -mc path/to/model.pdb -v -o model_based_locscale.mrc
The output will be a locally sharpened map scaled according to the refined atomic B-factor distribution of the supplied atomic model.

Recommended use of unfiltered input maps

Note that using unfiltered maps as input is essential. If using previously filtered maps, information beyond the spatial filter cutoff cannot be recovered.

Point group symmetry

If your map has point group symmetry, you need to specify the symmetry to force a symmetrised reference map for scaling. You can do this by specifying the required point group symmetry using the -sym/--symmetry flag, e.g. for D2:

locscale -hm path/to/halfmap1.mrc path/to/halfmap2.mrc -mc path/to/model.pdb -v -sym D2 -o model_based_locscale.mrc

Speed-up computation on multiple CPUs

To speed up computation, you can use multiple CPUs if available. LocScale uses OpenMPI/mpi4py for parallelisation, which should have been automatically set up during installation. You can run it as follows:

mpirun -np 4 locscale -hm path/to/halfmap1.mrc path/to/halfmap2.mrc -mc path/to/model.pdb -v -o model_based_locscale.mrc -mpi
If use of OpenMPI is not possible on your system, you can still take advantage of multiple CPU cores by using joblib. In this case, simply specify the number of CPU cores using the -np flag as follows:

locscale -hm path/to/halfmap1.mrc path/to/halfmap2.mrc -mc path/to/model.pdb -np 4 -v -o model_based_locscale.mrc