You are here

building a protein-dna homology model via cm_rosetta

5 posts / 0 new
Last post
building a protein-dna homology model via cm_rosetta

Hello everyone, recently I've been trying to make a homology model with a protein chain along with a nucleotide chain within it based on an existing template, but .

post-partial threading Rosetta dislocates the nucleotides nearly 1000 angstrom away from the structure. .

In principal, I'm suspecting this to be a result of an unwanted interaction between the nucleotide parameters and the protein which makes me wonder if their is a way to make cm_rosetta treat a specific chain/s as a rigid body ? In a sense, like I would with a ligand ?

My flags are these :

-in:file:fasta 1.fasta

-nstruct 20

-parser:protocol rosetta_cm.xml



-relax:jump_move true

-default_max_cycles 200



-hybridize:stage1_probability 1.0


Thanks in advance to all the helpers !


Post Situation: 
Wed, 2021-10-20 23:11

As nobody has answered I thought I'd give my two penny's worth.

The movement is the hybridiser's doing, not clashes from Lenard-Jones r^12 term —which I assume cut off after a given distance.

There are a few threading movers in Rosetta, including ThreadingMover which fills gaps if asked to do. I have not used with it scripts, only with PyRosetta (example):

threader = pyrosetta.rosetta.protocols.comparative_modeling.ThreadingMover(align=align[1], target_pose=target_pose, template_pose=template_pose)
threader.randomize_loop_coords(True) # default

I am sure it would work in scripts too...

Sat, 2021-10-23 10:30

Thanks matteoferla, i tried applying the simplethread mover to my model but I can't seem to get the xml format right. I also tried looking online for an example of it but couldn't find it, perhaps you can refer me to a relevant example of an xml file that applies it ?

Thanks in advance,


Sun, 2021-10-24 04:46

As mentioned, I don't know how to use Rosetta Scripts, so am of little help...

  • tests/integration normally contains a few life-saving XML scripts, but not for this it seems...
  • "SimpleThreadingMover" is in the "simple_moves" namespace, while "ThreadingMover" is in the "comparative_modeling" namespace (RosettaCM), so are different. The major difference is the latter runs off a Grishin file, as you've dealt with already for the "PartialThreadingMover", while the former just mutates the residues without dealing with indels —I am guessing the issues stem from that?
  • I should say that the fragment library part is optional and the accepted argument in PyRosetta is simply a vector1 of fragsets, which I am not sure how one does it in Scripts...
  • If you are keen on a mover and you are certain your problem is advanced... there is always the email addresses of the developers in the relavant header or source files...


Mon, 2021-10-25 03:17

thanks a lot for your response matteoferla, I believe I have resolved it using the backbone motions flag. so I consider this topic closed.

All the best,

Tue, 2021-11-02 01:47