How To Differentiate Protein Partners Between Homolog Protein Complexes

Question

[Reyhaneh]

I have Blast A against PDB and selected all the complexes. And the results are :

complex1: A1 - B
complex2: A2 - C
complex3: A3-D
……

where A1,A2,A3 are all blasts of A and B,C,D are the partners in the complexes. Using these complexes I have detected a binding site(BS) on the A.

Now I have selected a known PDB structure (A'-L') where A' is the homolog of A and L' is the partner. [In the calculation above for the binding site I have left out A'-L' complex]. But there is a big difference between the BS of A' -L' and the BS defined above.

It seems that L' is interacting with A' in a totally different manner, while all the other blast complexes of A show similar BS pattern when interacting with partners. Now I think there must be something different between L' and other partners (B,C,D), so my aim is to have a metric to compare these partners with each other and see how different one complex is to others.

Do you have any suggestion on how I can do this?

Thanks in advance;

[Neilfws]

"I would like to one single chain protein with these partners" does not make sense: please clarify. Also, be careful with the term "homolog" - it has a specific meaning which is not "similar". Also, "nature of proteins" is unclear - it's hard to define a metric without a sound quantitative measure. Can you be more precise about what you want to do?

[Reyhaneh]

I have changed the question and described more, Please tell me if there any ambiguous parts I need to edit. Thanks

[Reyhaneh]

I have changed the question and described more, Please tell me if there are any ambiguous parts I need to edit. Thanks

Answer 1

[DG]

So I am assuming since you are asking that there is no Crystal structure for the complex of A'-L' but there are for the other complexes. In that case there really is no way to compare A'-L' to the other complexes, especially if you suspect that the binding is occurring at a different binding site. To figure out A'-L' you are going to need to start looking in to binding site prediction algorithms and doing some molecular modelling of the proposed interaction to really sort things out.

[Reyhaneh]

Hi Dan; there is also an available complex for A'-L' in PDB. the fact that all A homologs apart from A' binds to their partner makes A'-L' complex a different case. Now I want to know what is especial about L' in comparison to other partners (B,C,D) that it makes A' to bind to it in different way. I hope i have clarified the problem.

[Reyhaneh]

Hi Dan; there is also an available complex for A'-L' in PDB. the fact that all A homologs apart from A' binds to their partner from the same site makes A'-L' complex a different case. Now I want to know what is especial about L' in comparison to other partners (B,C,D) that it makes A' to bind to it in different way (different from other available homolog complexes). I hope i have clarified the problem

[DG]

If you have the other complex as a crystal structure than I'm not sure what you are really asking. You should have a fairly good estimate of the binding sites now as well as the sequences that encode those binding sites. How exactly are A and A' related, are they paralogs or orthologs? What metric you want to use to measure how they are different depends on the question you are asking. For instance it may be that if they are paralogs that A has some transient interactions with L in the same manner as A'-L' that already existed. After duplication that interaction was refined in A'