You are here

Classic Relax

3 posts / 0 new
Last post
Classic Relax

Hello all,

I am currently working with the crystal structure of the ARAC protein domain and need a better understanding of the built in Classic Relax function. 

The ARAC domain consists of two seemingly identical chains that are very close and have strong interactions between each other. They naturally come with two, 10-residue long, "floppy ends" on the end of each chain that do not show up on the crystal structure. I have manually added these in via pymol, and after adding these in, the scorefxn is quite high and quite positive. Per advice from a colleague, I ran the ClassicRelax() function on the structure, and the scorefxn is now at a happy negative number. However, the structure of the two chains did change a bit. All the beta sheets and the alpha helices stayed intact, the two chains are slightly further apart now but they're still interacting. The floppy ends wrapped in different directions, with one curling in between the two chains. The chains are no longer symmetrical, and the overall structure has shifted somewhat.

I need to add a linker to this protein. Would it be best to go with the pre-relaxed structure or post-relax?    

Post Situation: 
Fri, 2019-06-28 09:24

Classic Relax has been deprecated for quite some time.  I would use FastRelax instead. As for working with the protein in subsequent Rosetta work, yes, a Relaxed structure is the way to go.  However, you don't want to relax too much or else you might not be able to get over local energy wells.  I suggest using the pareto-optimal method, which is basically FastRelax with coordinate constraints - Rosetta will be happy with the structure, the RMSD is very low, and it works as a great place to start.

Fri, 2019-06-28 13:34

iGemBoulder wrote:

> The chains are no longer symmetrical,

Also, just to add to Jared's answer-  if you want to preserve the symmetry, Rosetta has machinery to enforce that. You can start reading here:


Wed, 2019-07-17 12:43