Coral P2 Symmetry contacts

Interactive modelling (MASSHA, SASpy) and global minimization programs (SASREF, BUNCH, CORAL, GLOBSYMM)
Message
Author
M Wilce
Member
Posts: 4
Joined: 2009.09.26 05:58
Location: Monash University, Australia

Coral P2 Symmetry contacts

#1 Post by M Wilce » 2011.10.20 04:01

Is it possible to include contacts between symmetry related molecules in CORAL in the same manner as is possible in SASREF?

User avatar
AL
Administrator
Posts: 632
Joined: 2007.08.03 18:55
Location: EMBL Hamburg, Germany
Contact:

#2 Post by AL » 2011.11.01 13:30

Sure. The contact conditions file has a format similar to that of SASREF with the only difference that it refers to the chains instead of subunits(domains).

jcopps
Member
Posts: 6
Joined: 2008.12.31 02:01

Re: Coral P2 Symmetry contacts

#3 Post by jcopps » 2013.01.15 05:06

How does one specify contacts between P2 symmetry mates in CORAL? If I have a dimer and I want to create a contact of 2A between corresponding residues (let's say Leu1), do I write:

dist 2
1 1 1 1 1 1

or does the symmetry mate become its own chain, thus:

dist 2
1 1 1 2 1 1

Thank you

jcopps
Member
Posts: 6
Joined: 2008.12.31 02:01

Re: Coral P2 Symmetry contacts

#4 Post by jcopps » 2013.01.15 21:30

As a followup question, I've noticed that CORAL does not seem to correctly identify separate chains from my conditions file, which is written in the format I specified in my previous post. I'm dealing with a two chain system with P2 symmetry (so four chains total), and I'm trying to specify contacts between symmetry mates. My conditions file is

dist 17
1 6 6 1 6 6
dist 19
1 37 37 1 37 37
dist 14
1 101 101 1 101 101
dist 15
1 201 201 1 201 201
dist 16
2 241 241 2 241 241
dist 28
2 259 259 2 259 259

but CORAL will still identify all chains as "chain #1," and residues 241 and 259 of chain 2 will be identified by CORAL as 511 and 529 of chain 1, respectively.
Is this some kind of glitch, similar to the way CORAL identifies all residues as Gly residues in the condition files?

Thanks for any help that can be offered.

SaxsMax
Administrator
Posts: 342
Joined: 2007.08.03 19:16

Re: Coral P2 Symmetry contacts

#5 Post by SaxsMax » 2013.01.25 14:03

If I have a dimer and I want to create a contact of 2A between corresponding residues (let's say Leu1), do I write:
...
or does the symmetry mate become its own chain, thus:
dist 2
1 1 1 2 1 1
This is the correct way to specify the contact

emded
Active member
Posts: 28
Joined: 2014.07.04 11:34

Re: Coral P2 Symmetry contacts

#6 Post by emded » 2014.07.08 17:01

I have a similar problem. When I write this file it cannot be recognized by Coral. In addition, I don't know if the number of the residue needs to be the number specified in the PDB file or the number from the end of the N-terminus. As an example I have cut the protein into two parts so that Coral can build a loop in between. Therefore the chain two starts at residue 124 and not at residue 1. If one wants the first residue of the second chain to be used in the contacts, does it has to be written as number 124 in the restrain file or as number 1?

In addition,the state is a dimer of the protein with known dimerization interface from a crystal structure. My question is how I can use Coral to fit that dimerization interface but still allow to move the two monomers relative to each other in order to fit the structure better to the experimental data. As this interface is very small, there could easy be flexibility and movement of these two monomers relative to each other. Is the Coral the best program for such an analysis? I want both to add the missing loop in each monomer (which I know is flexible) and in addition I want to look at the movements of the two monomers of the dimer relative to each other.

Best regards,
Emil

SaxsMax
Administrator
Posts: 342
Joined: 2007.08.03 19:16

Re: Coral P2 Symmetry contacts

#7 Post by SaxsMax » 2014.07.08 17:27

I don't know if the number of the residue needs to be the number specified in the PDB file or the number from the end of the N-terminus. As an example I have cut the protein into two parts so that Coral can build a loop in between. Therefore the chain two starts at residue 124 and not at residue 1. If one wants the first residue of the second chain to be used in the contacts, does it has to be written as number 124 in the restrain file or as number 1?
Two parts connected by a linker is considered by Coral as one chain. So the number should be 124
In addition,the state is a dimer of the protein with known dimerization interface from a crystal structure. My question is how I can use Coral to fit that dimerization interface but still allow to move the two monomers relative to each other in order to fit the structure better to the experimental data. As this interface is very small, there could easy be flexibility and movement of these two monomers relative to each other. Is the Coral the best program for such an analysis? I want both to add the missing loop in each monomer (which I know is flexible) and in addition I want to look at the movements of the two monomers of the dimer relative to each other.
I would start with fixation of dimerization domains of the two monomers and modelling the position of the other domain(s)/ linkers.
If it does not allow to get a good fit then the dimerization domains can be unfixed while applying contacts restraint(s)

emded
Active member
Posts: 28
Joined: 2014.07.04 11:34

Re: Coral P2 Symmetry contacts

#8 Post by emded » 2014.07.08 17:49

OK, so I make a file called restrain.cnd which is a txt file and write:

dist 5.8
1 34 34 2 200 200

When Coral asks me for this file i enter restrain.cnd nothing happens but when i try restrain.cnd.txt than somethinng happens.

This is to restrain this position of my chain 1(a1.pdb) to my chain two (a2.pdb). This is so to say the monomer domain split up in two parts to allow Coral to build the missing loop from crystal structure. Chain one starts at residue 4 (first three are missing) and chain two starts at residue 124 as mentioned.

But the message I get is:

--EE-- Wrong format or conditions file
3 34 34 2 200 200.

I wonder what is wrong and why it can't read the file.

emded
Active member
Posts: 28
Joined: 2014.07.04 11:34

Re: Coral P2 Symmetry contacts

#9 Post by emded » 2014.07.08 17:51

Also the loop Coral makes without me performing the restrains is not connected to the chain, there is a small gap of about a residue, is this as it should be?

SaxsMax
Administrator
Posts: 342
Joined: 2007.08.03 19:16

Re: Coral P2 Symmetry contacts

#10 Post by SaxsMax » 2014.07.08 17:55

It should be ~3.8 A between the DR and the Ca atom in the structure

emded
Active member
Posts: 28
Joined: 2014.07.04 11:34

Re: Coral P2 Symmetry contacts

#11 Post by emded » 2014.07.08 18:19

SaxsMax wrote:It should be ~3.8 A between the DR and the Ca atom in the structure
Thanks for the reply, I think it is correct what I get, there is a gap of that size. Any suggestions why restrain file is not working? do you usually make the file in TextEdit? It worked like this for the config.con file for making the loop (it was also config.con.txt), however I have tried both restrain.cnd and restrain.cnd.txt file for the restraints and it won't read it.

I have tried with fixed chains and no restraints and I get a model for the loop that make the data fit much better (chi is 1.8 compared to 3.5 without the loop) however it would also be nice to try without fixing the domains and instead with restraining contacts but not the dimerization interface in order to see if there is movement between the two monomers.

SaxsMax
Administrator
Posts: 342
Joined: 2007.08.03 19:16

Re: Coral P2 Symmetry contacts

#12 Post by SaxsMax » 2014.07.09 10:13

The name restrain.cnd should work.
What exactly is the error message?

SaxsMax
Administrator
Posts: 342
Joined: 2007.08.03 19:16

Re: Coral P2 Symmetry contacts

#13 Post by SaxsMax » 2014.07.09 11:03

emded wrote:
Any suggestions why restrain file is not working?
Maybe you could actually post the content of the *.con and *cnd files
and also specify the number of Ca atoms in each domain (individual pdb file).
This would make the diagnostics easier

emded
Active member
Posts: 28
Joined: 2014.07.04 11:34

Re: Coral P2 Symmetry contacts

#14 Post by emded » 2014.07.09 20:07

SaxsMax wrote:
emded wrote:
Any suggestions why restrain file is not working?
Maybe you could actually post the content of the *.con and *cnd files
and also specify the number of Ca atoms in each domain (individual pdb file).
This would make the diagnostics easier
restrain.cnd file:

dist 7.0
1 53 53 3 56 56
dist 7.0
1 56 56 3 53 53

This is the symmetrical dimer interface I am trying to restrain. I have also tried with a single line for a single contact but it also doesn't work.

the content of the .con file is:

a1.pdb
LINK 22
a2.pdb
b1.pdb
LINK 22
b2.pdb

a1.pdb is composed of residues 4-99 and a2.pdb of residues 124-233 and numbers are same for b1 and b2.

I have a one domain protein which is a dimer with known crystal structure but an internal loop of 22 residues is missing. So I have split up the single PDB file into four (a1, a2 for one monomer and b1, b2 for the second monomer) so that the ends between a1 and a2 is exactly where the loop starts and ends.

I can get it to work without any restrain file, that is when I fix the subunits. However, I could not get it to work when I do not fix them but using the restraints. The error message is:

"--EE-- Wrong format or conditions file
1 53 53 3 56 56"

SaxsMax
Administrator
Posts: 342
Joined: 2007.08.03 19:16

Re: Coral P2 Symmetry contacts

#15 Post by SaxsMax » 2014.07.10 11:02

emded wrote: a1.pdb
LINK 22
a2.pdb
b1.pdb
LINK 22
b2.pdb
This means that there are two chains:
a1.pdb-LINK 22-a2.pdb and
b1.pdb-LINK 22-b2.pdb

Therefore, contact restraint should look like

dist 7.0
1 53 53 2 56 56
dist 7.0
1 56 56 2 53 53

Post Reply