Databases With Proteins Whose Binding Site Depends Strongly On The Ligand

Question

[Flow]

I would like to test the ability of some docking programs and my own one to predict the binding site of different ligands on the same protein. For that purpose I was reading literature in order to find examples of proteins where depending on the ligand, it bind in one part or another of the protein. But I found nothing relevant at the moment. I wonder if you know such databases or can you point to some examples from literature or wherever

[Flow]

can you argue why "exact duplicate"? I do not think so; the other question refers to "correlation between predicted and experiemental data" in general docking while my question here does not talk about correlation, just trying to find raw experimental data about proteins with different binding sites depending on the ligand. If you can tell me why both questions are the same and explain it to me, plase proceed so I learn new things I did not know before

[Flow]

another suggestions; rename biostar to sequence alignment stack

[Flow]

you can also suggest me where could I post such questions

[Michael]

I don't think it is an exact duplicate, the question is surely related but it's not frowned upon to ask related questions. Therefore, I'm willing to open it again, will do so unless some good explanation why this question is an exact duplicate is given.

[Flow]

thanks to everybody for the kind support !!!

[Michael]

So what about this question?

[Khader Shameer]

@Michael, thanks for taking this up. It should be possible duplicate - not exact duplicate.

[Khader Shameer]

@flow you have asked a question that deals with similar aspects here. Being someone who worked on structural bioinformatics / docking - I know thought they are pretty related and this question could be integrated into your first question.

[Khader Shameer]

@flow you have asked a question that deals with similar aspects here: Correlation Between Predicted And Experimental Binding Affinities In Docking, Good Examples?. AFAIK, they are pretty related and this question could be integrated into your first question. You are always asking related / possible questions - another example Which Are The Most Frequent Ligands In The Pdb Database? and Which Downloadable Database Of Pdb Ligand Binding Sites Would You Recommend?

[Flow]

If my questions are so similar, so should be also lots of questions posted here, but I appreciate your point of view

[Khader Shameer]

Yes @flow, I agreed many questions are similar. But if they are too close, you may combine them for better answers. This question is now re-opened.

Answer 1

[Khader Shameer]

I think you are looking for datasets for docking benchmark.

In this case, I can't give a specific answer without the details about your docking program. If I were you, I would look at recent manuscript that describe docking programs similar to your program and use one benchmark data from them.

Few pointers:

• If your docking program deals with protein-small molecule docking

See: http://zlab.bu.edu/zdock/benchmark.shtml ZDOCK is one of my favorite docking program. I think this dataset should be a good start. * You can also look in to co-crystallized protein-ligand complexes. For example browse PDB for tyrosine kinases co-crystallized with peptide substrates... etc.

• If your docking program deals with protein-DNA docking:

See: http://haddock.chem.uu.nl/dna/benchmark.html

• If you docking program deals with protein-protein docking

See: Critical Assessment of PRediction of Interactions. Also see the manuscript on the latest Protein-protein docking benchmark version 4.0.