EOM with a putative dimer of IDP

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
FrankG
Member
Posts: 8
Joined: 2017.05.17 11:02
Location: Lyon

EOM with a putative dimer of IDP

#1 Post by FrankG » 2018.01.03 19:46

Hi everybody,

I've recently obtained pretty good SAXS data about a near 40 kDa fully disordered protein (373 AA).
The "problem" is that the measured Rg is about 8.0 nm compared to the 5.6 nm calculated from the Flory's equation.

Then I have calculated the Rg for a theoretical IDP with the same sequence X2 (746 AA) and obtained 8.2 nm. I ran EOM with this sequence and the final fit is great so I imagine a dimer formed via the extremities of the protein.

My question is: how to configure EOM considering a dimer of IDP ?

Thank you,

Frank

mawhite
Active member
Posts: 20
Joined: 2011.08.15 22:18
Location: UTMB, Galveston, TX

Re: EOM with a putative dimer of IDP

#2 Post by mawhite » 2018.01.11 23:39

Frank,

When a symmetry interface is known, say in P2, then a distance restraint can be used in the Contacts file. In the case of IDP, you do not expect a symmetric dimer, so the Overall Symmetry should be set to [A]symmetry.

So how best to model your interface. Do you have (1) a core beta sheet interaction, which can be placed at the origin to fix the P2 symmetry, (2) a small dimerization domain, determined by homology, or (3) is its structure completely unknown, in which case just give a residue range in the Contacts file.

If residues (1-5) interact with the symmetry residues (15-20), antiparallel interface, this should provide correct restraints.
Contacts.cnd

Code: Select all

 dist 7.0
1   1   5 
1   15  20 
PS. Remember the quality of the results from EOM depend upon the correctness of the model used. An IDP model will give a poor fit if there is significant globular structure. EOM with a completely IDP chain will produce few models that are as compact as a globular domain in addition to an IDP region.

Best regards,
Mark

FrankG
Member
Posts: 8
Joined: 2017.05.17 11:02
Location: Lyon

Re: EOM with a putative dimer of IDP

#3 Post by FrankG » 2018.01.16 16:44

Dear Mark,

Thank you for your help.

My protein is predicted as 86% disordered by metaPrDOS and no homologous protein is available ... Certainly the few ordered residues are made by beta-sheets according to circular dichroïsm experiments but how to know if they are responsible for the interaction ? I wondered by adding Cl.Gu and try new SAXS experiment hoping to completely unfold the protein and see if the Rg change drastically to a monomer ...

Thank you for the contact file scheme but how do you know the distance to input ?

Best Regards,

Frank

mawhite
Active member
Posts: 20
Joined: 2011.08.15 22:18
Location: UTMB, Galveston, TX

Re: EOM with a putative dimer of IDP

#4 Post by mawhite » 2018.01.16 17:51

Frank,

The distance will be the minimum distance for the closest residue pair, by symmetry, within each specified range. Too close and there will be clash issues, too far and you are no longer specifying an interface. Since this is going to be a crude IDP dimer model, 6-8 A should be a safe choice.

PS. You may also consider using CORAL, or BUNCH, with a P2 dimer model to see what an 'average' model looks like. It would only represent an average of the ensemble and not have any real significance, but it is easier than EOM, and could help with the building of a suitable EOM model. In CORAL you can start without any distance restraints. Just remember not to over interpret any models that are produced.

Good luck.
Mark

FrankG
Member
Posts: 8
Joined: 2017.05.17 11:02
Location: Lyon

Re: EOM with a putative dimer of IDP

#5 Post by FrankG » 2018.01.16 19:30

Thank you very much for your help, I will try these different options.

Best Regards,

Frank

nmeri
Member
Posts: 4
Joined: 2018.04.06 01:10

Re: EOM with a putative dimer of IDP

#6 Post by nmeri » 2018.06.07 03:59

mawhite wrote:Frank,

When a symmetry interface is known, say in P2, then a distance restraint can be used in the Contacts file. In the case of IDP, you do not expect a symmetric dimer, so the Overall Symmetry should be set to [A]symmetry.

So how best to model your interface. Do you have (1) a core beta sheet interaction, which can be placed at the origin to fix the P2 symmetry, (2) a small dimerization domain, determined by homology, or (3) is its structure completely unknown, in which case just give a residue range in the Contacts file.

If residues (1-5) interact with the symmetry residues (15-20), antiparallel interface, this should provide correct restraints.
Contacts.cnd

Code: Select all

 dist 7.0
1   1   5 
1   15  20 


PS. Remember the quality of the results from EOM depend upon the correctness of the model used. An IDP model will give a poor fit if there is significant globular structure. EOM with a completely IDP chain will produce few models that are as compact as a globular domain in addition to an IDP region.

Best regards,
Mark

This may be a stupid/trivial question but I will shoot it here as I have asked a similar question in the main thread and did not get so much of a reply.
I am trying to define contacts between subunits of an IDP, fully disordered and no subunit pdb structure. I define core symmetry p2 and 1 domain. However, I cant get past the PDB file step of EOM. I tried creating a contacts file, with extension .txt and put that as pdb file (i know it is stupid but there isnt much help on the manual on this) and it didnt work. I also tried renaming the extension ".cnd" as you did but this time it complained that this isnt a viable pdb file.

Does anyone have a solution for this? I appreciate it.
Irem

Post Reply