You are here

suppress pdb output from backrub

8 posts / 0 new
Last post
suppress pdb output from backrub

Is it possible to suppress PDB output, but use silent output instead for the Backrub program? Rosetta <strikethrough>3.4</strikethrough> is used.

update: I am running 3.5, not 3.4.

Post Situation: 
Mon, 2014-10-27 11:16

I don't have a copy as old as 3.4 to test against, but try -out:file:silent <desired.filename>.

Mon, 2014-10-27 12:56

yes, that was the first thing I tried. Indeed silent files were generated. But PDB files were not suppressed :(

Mon, 2014-10-27 13:24

It looks like backrub in Rosetta3.4 is PDB file output only.

In Rosetta3.5 and later it's been converted to the "jd2" system, so the -out:file:silent option should work there.

Mon, 2014-10-27 13:06

... and if -out:file:silent is specified, are "low" poses packed in the silent files, or "last" poses? Is there any documentation for that?

Tue, 2014-10-28 02:41

I can't see where backrub in Rosetta3.4 would ever look at the -out:file:silent flag. Rosetta3.5 and later would, and the silent file should give you the same thing as the low PDBs. (And to confirm, you'll always get the _low and _last PDB outputs with the backrub application, even with recent versions of Rosetta and the -out:file:silent.)

Are you sure you're running Rosetta3.4, and if so, what is your full commandline?

Wed, 2014-10-29 10:36

My bad, I was using 3.5. Perhaps that's why -out:file:silent worked for me.

So the questions are:

1. how can I suppress the output of PDB files? I currently can only write a script to remove them periodically.
2. why are **low** structures generated in the files, while in Kortemme own papers they always design the **last** structures?

Thu, 2014-10-30 02:05

1) It doesn't look like you can suppress the output of the low and last PDB structures -- even with the most recent version of Rosetta you'll get them output.
2) I don't know for certain, but I'm guessing that since MC protocols typically recover the lowest energy structure, that was what was chosen for the "representative" structure of the run when things were converted to the "JD2" system. It looks like the person who converted things is different from the person who originally wrote the protocol, so I'm guessing it was an issue of missed communication on the part of the person doing the conversion.

Mon, 2014-11-03 15:47