Back To Multistage Rosetta Scripts
You can convert a traditional rosetta script to a multistage rosetta script XML format using the following command.
The converted script will be printed to the console if the -job_definition_file
option is not provided.
See here for more information and an example.
multistage_rosetta_scripts.default.linuxgccrelease -convert -parser:protocol existing_script.xml -job_definition_file new_script.xml
You can also convert a multistage rosetta script to a traditional rosetta script by using
-revert
or -reverse_convert
(both do the same thing).
Rosetta will load in the multistage rosetta script using the -job_definition_file
option
and will print the conversion to the console or to the filename defined by parser:protocol
if that option is provided.
Rosetta will only move the data elements from <Common/>
to the new script.
(WARNING: This is experimental and currently tends to underestimate by a factor of roughly 2x)
It can be difficult to know how many archive nodes you need to ask for ahead of time. In order to help, you can run the following command and it will predict the total amount of RAM needed to be devoted to archive nodes. You can divide this total by the amount of RAM per CPU in your system to calculate the number of archive nodes you should ask for.
Alternatively, you can run with the archive on disk option.
If you use the archive on disk option, you can use this feature to convert archive files to pdb files if you want to take a look at a few results while the job is still running.
Say you are storing your results in a directory called archives
and for whatever reason you feel like looking at result_400_1
(the result naming convention might change by the time you read this).
You can run the following command and the file archives/result_400_1.pdb
will be created
multistage_rosetta_scripts.default.linuxgccrelease -mrs:unarchive archives/result_400_1
-mrs:xml_template
prints an XML template skeleton to console or to a file if the -job_definition_file
flag is provided.
$> multistage_rosetta_scripts.default.linuxgccrelease -mrs:xml_template
<JobDefinitionFile>
<Job>
<Input>
<PDB filename="TODO"/>
</Input>
</Job>
<Common>
<SCOREFXNS/>
<RESIDUE_SELECTORS/>
<TASKOPERATIONS/>
<FILTERS/>
<MOVERS/>
<PROTOCOLS>
<Stage num_runs_per_input_struct="TODO" total_num_results_to_keep="TODO">
<Sort filter="TODO"/>
</Stage>
</PROTOCOLS>
</Common>
</JobDefinitionFile>
As with traditional rosetta scripts, you can pass the -info
flag
followed by a mover, filter, task operation, or residue selector name and Rosetta will
print the XML schema for that element to console.
For example, running:
multistage_rosetta_scripts.default.linuxgccrelease -info Idealize
gives
INFORMATION ABOUT MOVER "Idealize":
DESCRIPTION:
Idealizes the bond lengths and angles of a pose. It then minimizes the pose in a stripped-down energy function in the presence of coordinate constraints
USAGE:
<Idealize atom_pair_constraint_weight=(real,"0.0") coordinate_constraint_weight=(real,"0.01") fast=(bool,"false") chainbreaks=(bool,"false") report_CA_rmsd=(bool,"true") ignore_residues_in_csts=(residue_number_cslist) impose_constraints=(bool,"true") constraints_only=(bool,"false") pos_list=(int_cslist) name=(string)>
</Idealize>
OPTIONS:
"Idealize" tag:
atom_pair_constraint_weight (real,"0.0"): Weight for atom pair constraints
coordinate_constraint_weight (real,"0.01"): Weight for coordinate constraints
fast (bool,"false"): Don't minimize or calculate stats after each idealization
chainbreaks (bool,"false"): Keep chainbreaks if they exist
report_CA_rmsd (bool,"true"): Report CA RMSD?
ignore_residues_in_csts (residue_number_cslist): Ignore these residues when applying constraints
impose_constraints (bool,"true"): Impose constraints on the pose?
constraints_only (bool,"false"): Only impose constraints and don't idealize
pos_list (int_cslist): List of positions to idealize
name (string): The name given to this instance.