You are here

Multistate design output

7 posts / 0 new
Last post
Multistate design output
#1

Hi all,

Sorry to trouble again.

I understand that the multistate design protocol outputs structures where the chains A and B will become AA, AB and BB.

However, I get the output of chains A and B even in the output files AA and BB.

May I find out how I could produce the output files AA and BB homodimers?

Is this something to do with my corr file?

Category: 
Post Situation: 
Wed, 2018-06-20 00:21
tong

I would guess it's one of two things:

1) you don't have it set up to ever evaluate the homodimers, so it never creates them and never prints them to file

2) we have a misunderstanding over what A and B mean relative to one another - this sentence "However, I get the output of chains A and B even in the output files AA and BB." makes me wonder if this is the case.

How many structures are you using as input to multistate?  They are monomers or dimers?  Multistate is not docking, it cannot convert monomers to dimers, it can only re-thread sequences onto new equivalent-length backbones.  If you are getting monomers out when you wanted dimers you needed to have started with dimers.  If you are getting the wrong dimers out - do you mean the wrong BACKBONES or the wrong SEQUENCES?

 

Wed, 2018-06-20 05:59
smlewis

Hi,

thanks for your reply.

I input a IgG homodimer for the design of heterodimers. I want to evaluate the mutations that will favour the formation of heterodimer over homodimers. The initial input file is therefore a homodimer - the initial IgG molecule. I have also input the monomer chain A and monomer chain B. Am I doing something wrong here?

My output files only consist heterodimer molecules, I do not see any homodimer molecules.

Do you have an idea on what I may be doing wrongly with the protocol?

Greatly appreciate your help on this.

Wed, 2018-06-20 17:45
tong

A full IgG has four chains.  Is your input two heavy chains (working on the CH3 interface) or a light chain plus a heavy chain (fab fragment)?  

Wed, 2018-06-20 18:55
smlewis

Sorry my bad. I meant the two heavy chains, working on the CH3 interface.

Wed, 2018-06-20 19:57
tong

I assume it is your "states" files - or put another way the homodimers do not occur in your fitness.daf file, so they are never evaluated and printed.  Are you looking at the "protocol_capture/multistate_apl" demo or something else?

Thu, 2018-06-21 11:21
smlewis

I was trying to run the protocol on a IgG Fc region, referencing the protocols that were named in the protocol capture. Seems that you were right, I named the wrong corr in my states file. I shall try to run the program again. Thanks for the help!

Sun, 2018-06-24 23:30
tong