Hi Everybody,im trying to fit a Protein- Structure into a Cryo-Em-Map. It's a large Map of a ribosome. I had the fololowing code:
from modeller import *
log.verbose() env = environ()
struct='RPS25.pdb' map='Head40s.mrc' resolution=7.3 box_size=200 apix=1.83 x=0; y=0; z=0 #origin steps=100
# Read in cryo-EM density map den = density(env, file=map, em_density_format='MRC', voxel_size=apix, resolution=resolution, em_map_size=box_size, density_type='GAUSS', px=x,py=y,pz=z)
# Fit the PDB file into the map by MC simulated annealing den.grid_search(em_density_format='MRC', num_structures=1, em_pdb_name=struct, chains_num=[1], start_type='ENTIRE', number_of_steps=steps, angular_step_size=40., temperature=120., best_docked_models=1, translate_type='RANDOM', em_fit_output_file='modem.log') To search the Entire Map I chose ENTIRE as start type in the de.grid_search. It works until starting point 99. Than I get the following Errormassage:
C:\Dokumente und Einstellungen\Jochen\Desktop\Arbeit\modeller\FitRPS25>mod9v5 mo dem_mc.py 'import site' failed; use -v for traceback Traceback (most recent call last): File "modem_mc.py", line 25, in ? em_fit_output_file='modem.log') File "C:\Programme\Modeller9v5\modlib\modeller\density.py", line 61, in grid_s earch em_pdb_name, chains_num) IOError: openf______E> Cannot open file RPS2_init_**.pdb: Invalid argument
Any Idea how to solve that problem?
Problem Number two is the Voxelsize. My map has Voxelsize 1.83 but the rusulting maps allways has Voxelsize 1
Thanks for your help
Jochen
Jochen Ismer wrote: > im trying to fit a Protein- Structure into a Cryo-Em-Map. It's a large > Map of a ribosome. I had the fololowing code: ... > To search the Entire Map I chose ENTIRE as start type in the > de.grid_search. It works until starting point 99. Than I get the > following Errormassage: > > C:\Dokumente und > Einstellungen\Jochen\Desktop\Arbeit\modeller\FitRPS25>mod9v5 mo > dem_mc.py > 'import site' failed; use -v for traceback > Traceback (most recent call last): > File "modem_mc.py", line 25, in ? > em_fit_output_file='modem.log') > File "C:\Programme\Modeller9v5\modlib\modeller\density.py", line 61, > in grid_s > earch > em_pdb_name, chains_num) > IOError: openf______E> Cannot open file RPS2_init_**.pdb: Invalid argument
Yes, there is a bug in the code - it uses only 2 digits for the starting point, so won't work for starting points > 99. This will be fixed in the next release, but for now you'll just have to use fewer than 100 starting points.
> Problem Number two is the Voxelsize. My map has Voxelsize 1.83 but the > rusulting maps allways has Voxelsize 1
I'm pretty sure the output maps should be generated using the same voxel size as the input map. So it may be that the voxel size is not being correctly written into the header of these maps. Are your output maps also MRC?
Ben Webb, Modeller Caretaker
Jochen Ismer wrote: ... > Problem Number two is the Voxelsize. My [EM] map has Voxelsize 1.83 but the > rusulting maps allways has Voxelsize 1
What is actually happening is that while MRC maps written out by Modeller have the same map size (nx,ny,nz in the header) as the original input file, the cell size (xlen,ylen,zlen in the header) is always written out as zero. Thus any other program that reads a Modeller MRC map and tries to determine the voxel size by calculating xlen/nx will get zero. Since this is obviously nonsensical, I'm assuming that whatever program you're using defaults to a voxel size of 1.0 in this case.
I have fixed this bug in Modeller so that the correct cell size (and thus voxel size) is written in output MRC files. This fix will be in the next release, 9v6.
Ben Webb, Modeller Caretaker
participants (2)
-
Jochen Ismer -
Modeller Caretaker