You are here


2 posts / 0 new
Last post

Hi Everyone, I was wondering if anyone knows how to set path for saving output pdb files in Rosetta protein docking. I was using following command line to run protein docking, didn't found any saved output pdb files.

/path to/rosetta_2014.35.57232_bundle/main/source/bin/rosetta_scripts.default.linuxgccdebug @docking.options -parser:protocol docking_full.xml -database /path to/database/ -out:suffix _full -nstruct 50 > docking_full.log -restore_pre_talaris_2013_behavior

Docking.option file

-docking # the docking option group
-partners AB_HL # set rigid body docking partners
-dock_pert 3 8 # set coarse perturbation paramaters (degrees and angstroms)
-dock_mcm_trans_magnitude 0.1 # refinement translational perturbation
-dock_mcm_rot_magnitude 5.0 # refinement rotational perturbation
-s 3gbm_HA_3gbn_Ab.pdb # input model
-run:max_retry_job 10 # if the mover fails, retry 50 times
-use_input_sc # add the side chains from the input pdb to the rotamer library
-ex1 # increase rotamer bins to include mean +- 1 standard deviation
-ex2 # increase rotamer bins to include mean +- 2 standard deviations
-corrections:score:use_bicubic_interpolation true # use bicubic interpolation of the phi/psi bins
-out # out option group
-file # out:file option group
-scorefile docking.fasc # the name of the model score file
-silent docking.silent # the name of the model silent file
-silent_print_all_score_headers true # print score headers for each model in the silent file

Docking.full.xml file

<InitializeFromCommandline name=ifcl/>
<RestrictToRepacking name=rtr />
<RestrictToInterfaceVector name=rtiv chain1_num=1,2 chain2_num=3,4 CB_dist_cutoff=10.0 nearby_atom_cutoff=5.5 vector_angle_cutoff=75 vector_dist_cutoff=9.0 />
<PreventResiduesFromRepacking name=prfrp residues=11,41,345 />
These movers are part of the minimization protocol put together in the minimize_iface ParsedProtocol
<PackRotamersMover name=soft_rp scorefxn=soft_rep task_operations=rtiv,rtr,ifcl,prfrp />
<PackRotamersMover name=rp scorefxn=score12 task_operations=rtiv,rtr,ifcl,prfrp />

<TaskAwareMinMover name=soft_min tolerance=0.001 task_operations=ifcl,rtiv type=dfpmin_armijo_nonmonotone chi=1 bb=1 jump=1 scorefxn=soft_rep />
<TaskAwareMinMover name=soft_min_sc tolerance=0.001 task_operations=ifcl,rtiv type=dfpmin_armijo_nonmonotone chi=1 bb=0 jump=0 scorefxn=soft_rep />
<TaskAwareMinMover name=min_sc tolerance=0.0001 task_operations=ifcl,rtiv type=dfpmin_armijo_nonmonotone chi=1 bb=0 jump=0 scorefxn=score12 />
<TaskAwareMinMover name=min tolerance=0.0001 task_operations=ifcl,rtiv type=dfpmin_armijo_nonmonotone chi=1 bb=1 jump=1 scorefxn=score12 />

This minimization protocol is adapted from Whitehead, Fleishman, ... , Baker. See the XML scripts available in the Supp Mat.
<ParsedProtocol name=minimize_interface mode=sequence >
<Add mover=soft_rp />
<Add mover=soft_min_sc />
<Add mover=soft_rp />
<Add mover=soft_min />
<Add mover=soft_rp />
<Add mover=min_sc />
<Add mover=soft_rp />
<Add mover=min />
<Add mover=rp />
<Add mover=min_sc />
<Add mover=rp />
<Add mover=min />

<Docking name=dock_low score_low=score_docking_low score_high=score12 fullatom=0 local_refine=0 optimize_fold_tree=1 conserve_foldtree=0 ignore_default_docking_task=0 design=0 task_operations=ifcl,prfrp jumps=1/>
<Docking name=dock_high score_low=score_docking_low score_high=score12 fullatom=1 local_refine=1 optimize_fold_tree=1 conserve_foldtree=0 design=0 task_operations=ifcl,prfrp jumps=1/>

<SaveAndRetrieveSidechains name=srsc allsc=0 /> Speeds the move from centroid to full atom mode
<Add mover=dock_low/>
<Add mover=srsc />
<Add mover=dock_high />
<Add mover=minimize_interface />

Post Situation: 
Tue, 2015-06-30 11:28

By default, output files are placed in the directory where you launched the Rosetta run. There's various flags which can control this, one of the big ones is -out:path:all, which takes the path where (most) output files will be written to. (There are certain outputs which ignore -out:path:all.)

In your case, the reason you're not seeing output PDBs is the

-out # out option group
-file # out:file option group
-silent docking.silent # the name of the model silent file

in your options file. This tells Rosetta to use the (Rosetta specific) "silent file" output format. (This is a single file which can hold multiple structures in a compact representation.) All of your structures should be there, encoded in the docking.silent output file.

If you want PDB outputs, just omit this line from your options.

Alternatively, you can keep the silent file output, and then use the extract_pdbs application to select certain models for PDB output. (use -in:file:silent to specify the file, and -in:file:tags to list the structures you want.) This works well when you have thousands of models generated, but only need a few of the best-scoring structures.

Fri, 2015-07-03 13:24