Backbone didn’t change in docking, why?

[Anonymous]

Hi everyone,
When I use DockingHighRes to dock a 9-mer peptide to a protein, I allow peptide backbone flexible by using MoveMap, but it didn’t change after docking, why? And how to make protein backbone changeable while docking?

code like this:
movemap = rosetta.MoveMap()
movemap.set_bb_true_range(start, end)
movemap.set_jump(jump, True)
highResDocking = rosetta.DockingHighRes(dockingMinScore, jump)
highResDocking.set_move_map(movemap)
highResDocking.set_min_type(‘dfpmin’)
# do some perturbation…
highResDocking.apply(pose)

Could some one help me? Or give some suggestion?

[Anonymous]

bump.

[Anonymous]

Sun,

Same thing for me. I added a MinMover to the docking procedure as a quick and dirty way to move my downstream protein around. This is probably not optimal, as far as sampling configuration space goes, and a better choice might be an MC backbone mover before the “slideIntoContact” method.

(Like the example from the tutorial…)

movemap = MoveMap()
for i in range(1,100):
movemap.set_bb(i,True)
movemap.set_chi(i,True)
minmoverfa = MinMover(movemap,scorefxn_high_min,’dfpmin’,0.5,True)

As far as the flexible docking code from the tutorials/workshops, I’m not sure why it’s not doing what it says it should do. movemap.show() after .set_bb_true_range(n,N) lists that the backbone along the given amino acid chain is flexible, however the Docking procedure is not utilizing it.

Has anyone else come across this issue?

[Anonymous]

I don’t think the docking movers you’re trying to use are meant to consider backbone flexibility. For what it’s worth, you’re doing exactly the RIGHT thing by throwing a minmover with minimizeable backbone movemap into the mix. If you want more sampling, throw in loop movers at the interface, I guess? There are publications pending on this sort of modeling but I don’t think anything is out for Rosetta3 or PyRosetta; DDMI was released for Rosetta2 (Jha et al).

[Anonymous]

Hi,
I’m using Rosetta3.1 API to write my own protocol now. In Rosetta3.1, we need to tell docking mover the loops range of pose, then the mover can perturb the downstream partner and sample backbone conformation of loops by a loopMover, followed repack and minimize the region we defined by movemap. So if we apply a docking mover to a pose without loops defined, the mover will not sample backbone conformation.

[Anonymous]

Sounds like it should work. I think the guy that did KIC is also developing a protocol to do that; I have also written a protocol to do that (and am currently writing the paper for it).

[Anonymous]

It’s docking with loop modeling, I have saw this in several papers.

[Author]

Posts