imp
imp copied to clipboard
Invalid RMF files produced (rigid body members marked as rigid, not nonrigid)
@pellarin reports that running https://github.com/salilab/mediator/tree/master/mediator_modeling/template/modeling.py generates RMF files that look like:
% rmf_show -v
...
+ "med3" [representation]
molecule (sequence): 1(s)
+ "Beads" [representation]
rigid body index (IMP): 262099(s)
translation: [29.2966, 3.14344, 55.0695]
rotation: [0.582091, 0.753277, -0.00997226, -0.306013]
- "med3_1-40_bead_floppy_body_rigid_body_member" [representation]
coordinates (physics): [0, 0, 0](s)
translation (physics): [10.0253, -74.7453, -30.4948](s)
rotation (physics): [1, 0, 0, 0](s)
Note that the translation and rotation of the floppy body are marked as static (s)
. This causes IMP to treat it as a rigid member when it reads it back in. This is incorrect because the floppy body is a nonrigid member.
@pellarin modeling.py
does not work with today's IMP, so cannot reproduce:
$ ~/imp/release/setup_environment.sh python modeling.py
autobuild_model: constructing med6 from pdb ../pdb_files/head_module_em_aligned_translated.pdb and chain G
autobuild_model: constructing fragment (1, 60) from pdb
autobuild_model: constructing fragment (61, 82) as a bead
autobuild_model: constructing fragment (83, 192) from pdb
autobuild_model: constructing fragment (193, 295) as a bead
Traceback (most recent call last):
File "modeling.py", line 79, in <module>
bm1.build_model(domains)
File "/clarinet1/home/ben/imp/release/lib/IMP/pmi/macros.py", line 459, in build_model
for k in super_rb:
TypeError: 'NoneType' object is not iterable
Latest IMP+PMI+mediator do not show this behavior for med3_1-40_bead_floppy_body_rigid_body_member
.