can remodel add loops in one chain in multichain PDB structure?

[Anonymous]

I’m trying to model 4 loops within each of 4 chains within one PDB file. Modelling the loops simultaneously is important because all chains (and also 4 another without any loops to model but interacting with that 4 remodelled) are assembled into oligomer, so that loop formation should take into account other chains. I have:

• renumbered PDB file, with all chains renumbered from 1 
• a blueprint for one chain with coordinates corresponding to residue numbers in the PDB file, separate for every chain – all the loops are closeable and all the definitions work fine individually and for an individual chain in an input PDB file (getBluePrintFromCoords.pl script takes one chain as option).

To begin, I tried to take 4 chain PDB file and add loops within one chain (remodel allows defining one blueprint without chain specification within blueprint, so that’s natural). However 

My command is:

$rosetta/remodel.mpi.linuxgccrelease -s A.pdb -database /opt/rosetta/main/database -remodel:blueprint chosing_c2_tm_clean_sort_AB_noPIKAA.remodel  -run::chain A -overwrite -nstruct 1

and the procedure fails to close the loops. So my question is: can remodel add loops in one chain in multichain PDB structure or should I look for some workaround, i.e. renaming the whole structure into one chain, renumber all residues from one and after all renumber residues and assign to separate chains back again?

[Anonymous]

I cannot answer your question, but I can say I encountered unique issues two years ago when I specified chain and I ended up simply making a large single chain protein as you suggest. Remodel is sporadically fussy with PDB vs. pose numbering, so I am not too surprised. About the renumbering I have a two-way python with pymol2 script for it, if that helps?

About your imput, for loops I find changing -generic_aa to A or G very helpful. Also the documentation also is religious about -ex1 -ex2, I simply obey.

[Author]

Posts