I am trying to do model a loop using loopmodeling mode and want to keep the rest structure rigid. when I run loopmodel.linuxgccrelease. A lot of WARNING lines like
core.io.pdb.file_data: [ WARNING ] discarding 3 atoms at position 17
in file . Best match rsd_type.
core.io.pdb.file_data: [ WARNING ] discarding 9 atoms at position 30 in file . Best match rsd_type
I found that some positions in other regions were changed after loop modeling.
I tried -loops:natsc and -loop:refine_only. seems both changes the side chains
It's probably balking because something is wrong with the sidechains at those positions - most likely either the hydrogens are labeled improperly (and ignored) or there are heavy atoms missing, or both. Rosetta won't tolerate missing atoms and will build them back in; it then repacks the sidechain to ensure the newly built atoms don't clash.
These changes are probably occuring at the time of PDB read-in, not within loop modeling (thus loops: natsc didn't work).
Try -pack::pack_missing_sidechains false. It defaults true so you need the false.
If the moving sidechains are far from your loop you they won't affect its conformation - you can also just fix them by editing the PDB later.
Seems it worked. BUT I always had a lot of WARNING lines like
discarding atoms at position in file Best match rsd_type
Is this step of discarding some atoms of each residue and best match rsd_type required?
It's not "required", in the sense that a perfectly formatted PDB will not produce these warnings. That said, it's probably nothing to worry about. You could try stripping all the hydrogens from your input file and see if that makes the warning go away.
After I further checked out. I still saw slight movement of sidechains of many residues. the positions of alpha C are same. what is the reason. thanks
-loops::frag_sizes 9 3
-loops::frag_files fragment9 fragment3