You are here

fragment-based loop modeling to get a helix at the loop site

12 posts / 0 new
Last post
fragment-based loop modeling to get a helix at the loop site

Hi every body,

I am a new Rosetta user,
I want to model loops of my protein by preforming using fragment-based loop modelling of Rosetta,(De Novo protein-designing module).
I will use this command for modelling
-loops:remodel quick_ccd
-out:prefix _

But I don’t know how can I find my fragments file from database and which database,
would you please help me in this subject.


Post Situation: 
Wed, 2013-05-29 06:00

To generate fragment files, we highly recommend you use the Robetta web server - - For most uses it produces perfectly acceptable fragments at a tiny fraction of the headache.

If you really need to do something complex or custom, keep in mind that there's an old way of picking fragments and a new way, and you'll be better off in the long run using the new fragment picker. (

Wed, 2013-05-29 11:09

If you know you have a helix in the loop, and have an idea about where it is, you could look into RosettaRemodel:

Mon, 2013-06-03 09:50

How can I use RosettaRemodel in Rosetta 3.4, I couldn’t use it,

Wed, 2013-06-05 09:17

It should be in Rosetta3.5.

Wed, 2013-06-05 09:38

But when I employ this command I did not get anything
/rosetta_tools/remodel/ -pdbfile Loop_0911.pdb

Wed, 2013-06-05 09:52

Does it give you any output at all?

One possibility is that you have a chain designation issue. By default it looks for chain "A" or chain "_". If your PDB doesn't have that chain (say it only has a single chain "H"), then the script won't output anything. Either change the chain designation in the PDB, or specify the chain on the commandline:

rosetta_tools/remodel/ -pdbfile Loop_0911.pdb -chain H > blueprint.txt

Wed, 2013-06-05 10:22

This is a pdb file that I generated by loop-model.

Wed, 2013-06-05 10:33

Hmmm. Do you get something that seems reasonable for other (closely related) PDB files? Would it be possible for you to attach the PDB file that's not working here? (You may need to change the extension to .txt for the forum software to accept it.) If the -chain option doesn't fix things, it's likely some edge case for that particular PDB.

Wed, 2013-06-05 15:06

I attach my PDB file please find,


Thu, 2013-06-06 00:19

Running: -pdbfile TM4_0001.txt -chain R > blueprint.txt

With the Rosetta3.5 version of gives me a blueprint.txt file that looks like:

1 N .
2 Y .
3 Y .
4 A .


This is pose numbered (sequential starting from one) rather than PDB numbered, -- As Jared and indicates, you'll need to renumber your PDB starting from 1. (Either the way Jared does, or by using the -out:file:renumber_pdb option on a Rosetta run. Using the score application with -out:pdb works well for this.)

It's not much of a blueprint file - you'll need to edit it according to to do the remodeling that you want, but it's certainly a starting point for your protocol.

Thu, 2013-06-06 13:15

Remodel requires renumbering from 1 as far as I know, so I used the PyRosetta Toolkit GUI ( ) to renumber your PDB and output a basic blueprint file you can edit in a text editor. I'll look into what is wrong with getBluePrintFromCoords. I never had any luck with it either.

Thu, 2013-06-06 07:51