You are here

Docking with constraint

3 posts / 0 new
Last post
Docking with constraint
#1

Hi everyone,
I am trying to dock a domain into a dimer protein. The domain is lowest score structure that was generated by abinition relax. Because they are continous amino acid sequence, so i use atompair constraint betwen c terminal of first monomer with n terminal of predicted domain. I also constraint its c terminal with n terminal of the second mononer.
The problem is the distance between 2 pair of Ca atom is too big to satisfy the constraint condition. The job can not return the out put file. Does anyone have experience about this issue. Would you please teach me how to slove this.
Thank you so much.
Hongtham

Category: 
Post Situation: 
Sat, 2014-12-27 09:56
Hongtham

I think I need solve this issue in Abinitio step. In this step I need to specify the position of C and N terminal of input sequence. And then the result structure can have the correspoding N and C terminal with the rest part of protein. But I dont know how to do it? Can any one help me?
furthermore. In my case, I have a amino acid sequence like this: partA - domain I - part B - domain II. PartA and part B I got structure by homology modeling (Swissmodel), but domain I, domain II, I had to run Abinito relax to get their structure. My job is assemble all these part into one which is partA and partB have been in specific position already. And now I am having trouble with above problem when I m trying to dock domain I into partA_PartB. can any one help me?
Thank you so much.
Hongtham

Sat, 2014-12-27 22:12
Hongtham

The first thing I'd do is filter your abinitio models to discard those where the termini are positioned such that they wouldn't be able to be joined to other domains (e.g. if they're buried slightly, etc.). The standard post-analysis methods are good for regular monomers, but if you have additional information (e.g. the N/C-terminal is connected with another large chunk) there's no problem incorporating that into your analysis/filtering. How to do the filtering is a little tricky, and depends on what you want out of the system. I'd recommend taking a look at the top models and seeing if it looks plausible that they form a connection. If not, drop the models from consideration.

Assuming you have models which you think could form a continuous chain, the question is why aren't they in the docking? The first possibility is that you have your CA constraints set too tight. Abinito is unlikely to put the termini in exactly the right location to form a covalent bond - you'll probably need to do loop remodeling to join them. Loosen your constraints to allow more movement and see if you get reasonable models that way.

Further than that, I'd need to know more about what sort of error messages you're getting, details on how you're running your protocol, etc.

Fri, 2015-01-02 09:43
rmoretti