Bunch with multiple chains

Linear (OLIGOMER), and non-linear (MIXTURE) analysis, singular value decomposition (SVDPLOT), addition of missing fragments (BUNCH, CORAL), analysis of flexible systems (EOM/RANCH & GAJOE), flexible refinement of high-resolution models (SREFLEX)
Post Reply
Message
Author
Nicholas
Member
Posts: 9
Joined: 2010.04.06 15:44
Location: Faculty of Life sciences, University of Copenhagen

Bunch with multiple chains

#1 Post by Nicholas » 2014.03.27 16:07

Hi,

I'm wondering if it is possible to model a molecule containing two chains using bunch?
I have a molecule consisting of chain A and chain B. Chain A together with a part of chain B (that I will call B1) form a domain with a structure defined in a pdb file. I'm not interested in the relation between chain A and B, this I already know fronm the pdb, what i'm interested in is the rest of chain B.

the problem is illutrated below:

-----CHAIN A------
-----CHAIN B1-----|--FLEXIBLE--|-----CHAIN B2----

CHAIN A, CHAIN B1 and CHAIN B2 are structured domains with a pdb file.
FLEXIBLE is unknown.

The question is how does A+B1 relate to B2 via the flexible part of chain B

I've tried to add the A sequence in front of the B seqence in order to cheat BUNCH into using it as one long chain. As I provide a PDB file for this region I thought that preBUNCH would just place beads at the Calphas, however it fails the sequence alignment (I have checked that the sequence in the merged A+B1 pdb file is the same as in the sequence file). I guess preBUNCH checks if the Calphas are placed close enough to form a chain and it therfore fails.

Is there some way I can cheat BUNCH into treating the molecule as one chain? Or should I try something else?

Best regards
Nicholas

Alex
Active member
Posts: 692
Joined: 2007.08.09 21:10
Location: Planet Earth

Re: Bunch with multiple chains

#2 Post by Alex » 2014.03.27 16:29

Hi Nicholas,

unless you have a very long flexible part (>100aa), you can also use CORAL to do this modeling by fixing chain A and B1.

Otherwise, the way you did it should have worked if you fused the sequence and put A-B into one PDB file.
Probably, you might want to remove TER/END fields too. If I recall it correctly - I did smth similar some time ago with BUNCH.

If i were you, i would also check if pre_bunch would work on just:
- B1 - B2
and
- chain A
to make sure there are no misalignment(s).

I guess preBUNCH checks if the Calphas are placed close enough to form a chain and it therfore fails.
i don't think so.

HTH,
Alex

Nicholas
Member
Posts: 9
Joined: 2010.04.06 15:44
Location: Faculty of Life sciences, University of Copenhagen

Re: Bunch with multiple chains

#3 Post by Nicholas » 2014.03.27 17:26

Dear Alex,

Thank you for the reply, so I was trying the right thing. I will have to go back and check by sequence again. :-(
But as you say, CORAL might be a better tool.

Best regards
Nicholas

Nicholas
Member
Posts: 9
Joined: 2010.04.06 15:44
Location: Faculty of Life sciences, University of Copenhagen

Re: Bunch with multiple chains

#4 Post by Nicholas » 2014.03.27 17:31

Just to update, if anyone has a similar problem. After reading the CORAL manual I see that CORAL is much more fit for this task.
Last edited by Nicholas on 2014.03.27 20:32, edited 1 time in total.

Alex
Active member
Posts: 692
Joined: 2007.08.09 21:10
Location: Planet Earth

Re: Bunch with multiple chains

#5 Post by Alex » 2014.03.27 17:46

Well it depends:
1. sampling of conformations to model is better in BUNCH than in CORAL
2. how flexible is your protein? do you have Kratky/dimensionless Kratky plots?
3. if it is flexible, you might wanna use EOM and to do so here u would need to cheat RANCH the same way
you tried to cheat BUNCH.

Nicholas
Member
Posts: 9
Joined: 2010.04.06 15:44
Location: Faculty of Life sciences, University of Copenhagen

Re: Bunch with multiple chains

#6 Post by Nicholas » 2014.03.27 21:03

The protein has a fairly defined structure, with a closed kratky plot.

I tried CORAL but it keeps moving some of the subunits, relative to each other, even when they are grouped.
Even when I explicitly specify that they should be fixed.

I also tried to fix all subunits, however I can see that the are still moved.

Does coral have some criteria where it will move the subunits no matter what. For example, if there is a steric clash.

Best regards
Nicholas

Alex
Active member
Posts: 692
Joined: 2007.08.09 21:10
Location: Planet Earth

Re: Bunch with multiple chains

#7 Post by Alex » 2014.03.28 00:52

it is hard to say without seeing the data and input PDBs.

It could be that your subunits are too far from each other or they are much off the origin.
Does coral have some criteria where it will move the subunits no matter what. For example, if there is a steric clash.
there is a set of penalties. if the penalty value is too high and can't be reduced, I think, it may move the subunit ( i am not 100% sure about it though). I would check your initial subunit positions first.

HTH,
Alex

Post Reply