The unofficial mini-rosetta tutorial
This page is a slowly growing collection of useful tips and trips for mini-rosetta. Our most recent build is typically on the Cassini cluster in /software/minirosetta/latest/.
To learn more about rosetta, visit the Rosetta Commons webpage.
Rosetta Options
Mini-rosetta is split into several binaries. The minirosetta binary can perform the following procols: ab initio modeling, ligand docking, relax, loop relax. It defaults to abrelax (ab initio) when no protocol is explicitly specified.
To optain a list of vailable run-time options, run: minirosetta -options:all
All runs will require the minirosetta_database. The location defaults to ~/minirosetta_database/. We have a copy on cassini in /software/minirosetta/minirosetta_database/. You can either create a symlink in your home directory to that location, or specify the location of the database at runtime with -in:database /software/minirosetta/minirosetta_database/
Design
Design is the process of (re)designing the amino acid sequence for a given staring structure. At the minimum, a starting structure and a resfile are required (fixed backbone mode) or a fragment library (flexible backbone). See under Ab Initio for info how to make a fragment library.
The Resfile
The resfile defines which residues should be designed and how. The format is a follows:
Coming soon.
Fixed Backbone Design
Coming soon.
Flexible Backbone Design
Coming soon.
Ab Initio
In ab initio modeling, a structure is assembled from short fragments from the PDB. The first step is to make a fragment library for a given sequence or secondary structure definition, then run rosetta to pick fragments from that library and assemble them into a model structure. The latter step is repeated many, many times generating many structures.
Making Fragments
There is a nicely adjusted make_fragment.pl script in /software/rosetta/nnmake. Run it without parameters to see available options. If making a fragment library for a given amino acid sequence, put the sequence in a file called TESTA.fasta and run:
/software/minirosetta/nnmake/make_fragments.tpl TESTA.fasta -nojufo -nosam
The process takes several minutes, depending on the length of the sequence.
Two library files will be generated, one with 3 residue fragments and one with 9 residue fragments. The files are aaTESTA03_05.200_v1_3 and aaTESTA09_05.200_v1_3.
Alternatively make_framents.tpl can build a fragment library from a secondary structure definition, one made by PSIPRED or one made by you from scratch. Use the -nopsipred and -psipredfile options. This option still requires a FASTA file with the sequence (can be all A's). To remove all sequence bias, nnmake++.exe (found in the same directory) needs to be re-run with the -ssonly flag.
Running Ab Inito mini-rosetta
The ab initio run requires the FASTA sequence and the matching fragments library generated above. A secondary structure definition is useful but not required. Specify all components on the command line as follows:
/software/minirosetta/latest/minirosetta \
-abinitio \
-out:file pdb \
-out:file:prefix OUT \
-in:path:database /software/rosetta/mini/minirosetta_database/ \
-in:file:fasta TESTA.fasta \
-in:file:frag9 aaTESTA09_05.200_v1_3 \
-in:file:frag3 aaTESTA03_05.200_v1_3
You may, of course, add additional arguments to specify the number of cycles, restraints, number of structures to build, etc. A single run takes several minutes, again, depending on the size of the model.
The generated model will be in OUT_0001.pdb. Details scores will added to score.fsc.
