I am struggling to generate a symmdef file for a helical model and I am hoping someone can help.
The helical parameters are as follows: rise = 17 Angstrom with 4.674 subunits per turn. The helix is apolar i.e. the ends of the helix are the same as related by 180 degree rotation. The repeating unit is a dimer with C2 symmetry. In the attached pdb (ca only to avoid upload limit) the A and B chains form the dimer which is related to the next C/D pair by the helical symmetry I mentioned.
I can generate a symmdef file and correct model for one of the monomers of the dimer using the following command:
make_symmdef_file.pl -m HELIX -a A -b C -p 4monomers_finetuned.pdb
This generates a helix comprising chain A, chain C … and 4 subunits in either direction. However, when I try to include the C2 symmetric neighbour of chain A using the “-I” flag it fails using this command:
make_symmdef_file.pl -m HELIX -a A -b C –i B -p 4monomers_finetuned.pdb
The result is a pdb with all the chains aligned along the helical axis in a nonsensical fashion i.e. overlapping subunits.
Is there something I am doing wrong?
Unfortunately, nonpolar helical symmetry currently unsupported with make_symmdef_file.pl.
One possible workaround would be to make A&B (and C&D) to be a single chain. I edited the input PDB you provided to make it two chains instead of four, and then used the command you gave, and the symmetric model looked quite reasonable. Depending on what you're doing Rosetta isn't too picky about "chains" being covalently connected, so pretending that the two chains are really one chain is probably not too much of an issue. (If you get issues, try adding -missing_density_to_jump to your Rosetta commandline options.)