There are a hand full of options, that are useful most of the time you run Rosetta. Here, we will briefly go through those.

1. Controlling input

Database (option normally not required)
-database /.../Rosetta/main/database

During a Rosetta run, a lot of information of pulled from the Rosetta database. It contains e.g. score files (weights for individual energy terms), bond lengths and atom connectivity for amino acids. By default, Rosetta finds it's database automatically, and this option is NOT required. If you have a non-standard intallation, specify the path to your local Rosetta database Following the -database flag.

Provide input structures

Here is how you pass structure files to be used as input. This can be a structure where you want to re-design the sequence, re-design a loop, or two subunits that you want to dock together.

-in:file:s <pdb file>  
-in:file:l <list of multiple pdbfiles>
-in:file:native <pdb file>

-in:file:s is the way to provide a single input structure. A list is a text file with pdb files (and their path, if located in a different directory). Rosetta will go through the files consecutively. The native flag gives you the opportunity to provide a reference structure for comparison. E.g if you have a guess how two proteins interact, you can compare to this structure or when re-designing the backbone of an existing protein.


This option is very important if Rosetta fails to read in your pdb file because of non-standard chemical entities in the file.

Provide input sequence

For ab initio structure predictions, you will have to provide a sequence in fasta format. Here is how you do that:

-in:file:fasta < e.g. my_sequence.fasta >

2. Controlling output

Change default file names

Rosetta lets you to modify the names and format of output files as well as the amount of information given to you.
Often you will see a structure file and a score file output. Structure files are named after the input file and automatically get numbers appended (e.g. MyFile_0001.pdb). You can change score file names and add a prefix or suffix to all output file names. All of these are optional.

-out:prefix < e.g. trial2_ >
-out:suffix < e.g. _trial2 >
-out:file:scorefile < e.g. abinitio_scores.dat >

The -overwrite option is needed if you already has a structure file (like MyFile_0001.pdb), and you re-run Rosetta with the same input file. In that case, Rosetta will fail if this option is missing.

Silent files - combine many output pdb files in a single compressed file

When running Rosetta, you will often get hundereds or thousands of individual pdb files. It can be useful to instead have a single file, where all structural information is combined in a binary file (mostly unreadable for humans). They take up less space, and let you extract only a small number of structures (e.g. best energy structures) after a run is finished. They get the .out file extension

-out:file:silent_struct_type binary
-out:file:silentfile < e.g. design_1.out >

Look at the Analysis chapter to find out how to extract pdb files from silent files.

Control the number of runs
-nstruct < e.g. 1000 >

Typically, you run a simulation in Rosetta multiple times. This is important, because the sampling in an individual run is normally much too limited. -nstruct 1000 will run 1000 iterations of a simulation and output 1000 files: