I have been running snugodck jobs on Rosie for a while but queue times are long so I am trying to run jobs locally. I am getting the error below:
ERROR: Sequence for partner1:
does not match first member of ensemble1:
Prepack runs fine using:
../main/source/bin/docking_prepack_protocol.macosclangrelease -in:file:s 136C_1.pdb -ex1 -ex2 -partners LH_EFG -ensemble1 antibody_ensemble.list -ensemble2 antigen_ensemble.list -docking:dock_rtmin
For snugdock I am using: (only 5 structures just to see if it is working):
../main/source/bin/snugdock.macosclangrelease -s 136C_1.pdb -ensemble1 antibody_ensemble.list -ensemble2 antigen_ensemble.list -antibody:auto_generate_kink_constraint -antibody:all_atom_mode_kink_constraint -partners LH_EFG -spin -dock_pert 3 8 -ex1 -ex2aro -nstruct 5
The error says that sequence for partner 1 is the heavy chain, but in the PDB files the light chain is first.
Can you check via a TextEditor (this is important because PyMOL will order chains alphabetically) whether or not the prepacked files listed in "antibody_ensemble.list" have the same order as the file "136C_1.pdb"?
Due Rosetta's internal representation, all the files have to have the chains in the same order and this order has to match the "-partners" flag.
I checked all the files by texteditor and they all are L followed by H. FYI - attached are a prepacked file from the antibody_ensemble.list, and also the partner1_starting_conf file, and 136C_1.pdb (I truncated the antigen part so that file is less than 512KB limit for uploading). (I put txt as the file extension since ppk cannot be uploaded).
136C_1.pdb is an output file from a snugdock run on the ROSIE server. 136C_1.pdb is fine as input when I use snugdock on the ROSIE server.
Thanks very much for your help.
First, sorry about the long wait. I was sick and then traveling for the holidays. Looking at these files, the command line you described should work, so I am surprised. I did not know about the upload limit on the forum. I would like to try to replicate the error on my machine. Would you mind emailing everything necessary to do so?
You can contact me at firstname.lastname@example.org.
Thanks very much. I will send you everything by email.
In the meantime I have been running snugdock with just a single antibody and antigen, without ensembles, on a local platform. I am trying to repeat a fast snugdock trial that I ran previously on Rosie, to make sure that I get similar results. I use the snugdock commands given above but without the ensemble commands. The runs finish but I find that the range for Irms is much larger than for the fast snugdock run on Rosie. What flags corresponds to fast snugdock on Rosie?