You are here

get_distances command in rosetta 3.5

4 posts / 0 new
Last post
get_distances command in rosetta 3.5

Hi Rosetta Support group,

I was wondering if Rosetta 3.5 version has get_distances command available? I looked into rosetta source/bin/ directory to locate it but could not. Could you please suggest where to find it in Rosetta 3.5 if it is included in this version.


Post Situation: 
Mon, 2014-01-20 11:31

I'm not familiar with any sort of "get_distances" command for Rosetta. What are you expecting it to do? (Or more precisely, what is the scientific problem you want to solve?) There may be an alternate way of obtaining the results you want.

Failing that, getting distances between arbitrary sets of atoms is relatively straightforward in PyRosetta, or it can also be accomplished with PyMol. If you're only interested in a few distances, we may be able to put something together with RosettaScripts , though that's painful if you're looking for a large number of them.

Mon, 2014-01-20 16:18

hi rmoretti,

Thanks for the response.
In our lab earlier with rosetta 3.1 on linux we used get_distances.static.linuxgccrelease command for calculating the distances between a particular ligand atom and specific protein atom in the docked complexes obtained from Rosettaligand docking. Now, we have latest version of Rosetta 3.5 on mac as well as linux, but i could not locate this command.

Basically, we want to see how far is a particular ligand atom from the specific protein residue atom in a set of docked complexes and select the best one based on this criteria. We have around 25000 docked complexes to analyze for this parameter.

Please suggest a fix for this issue.


Mon, 2014-01-20 19:21

Interesting. I can't seem to find a get_distances application in the version of the Rosetta3.1 release that I'm looking at - nor is it in the developer's version of Rosetta from around that time. Could this be a custom application that someone in your lab extended Rosetta with? If so, it should be simple enough to update it for 3.5. (Actually, if you still have the source directory for your 3.1 release hanging around, we could pull out the application and update it for 3.5; although I doubt much "updating" will be needed.)

In any case, since you're interested in a single distance between a particular atom on a specific residue and another specific atom on another specific residue, the "modern" way I would do it would be by using a RosettaScripts filter. Specifically the AtomicDistance filter ( You can specify the two residues and atoms which you're interested in looking at, and if you add "confidence=0" to the tag, you'll turn off the filtering aspect of the filter, and turn it into a reporter instead. Then just add that tag to the FILTERS section of the skeleton XML ( ) and add the filter to the PROTOCOLS block (you don't need the mover), and you can run your structures through the rosetta_scripts application with that XML, and your distance will show up in the scorefile under the same name as your filter. When you do this, you'll probably want to add the flags "-out:file:score_only" and "-no_nstruct_label" to the commandline. The first keeps Rosetta from outputing the (unchanged) structures, and just outputs the scorefile, and the second doesn't add the _0001 to the structure name, so the names in the output scorefile will match the input filenames/tags.

Tue, 2014-01-21 07:37