Proteins bound to more than one ligand

Question

[abbax]

I'm at the beginning with the bioinformatics. Until now I did only theory about this subject at the university, now I would like to start to do something practical.

In particular I would like to search, on the PDB, the structures in which proteins are bound to more than one ligand. In the advanced search of the rcsbPDB there is only the option "has ligands"; how could I delete the structures with only one ligand?

[abbax]

sorry for the multi post but nobody can help me?

[Michael]

Maybe you should give it some time, especially during the week-end and the football world-cup final (yeah!). Bookmark your question, and update it with your progress. A typical counter question to posts like yours would be "what have you tried?". If possible post some code you have written or found that almost does what you want. I recommend not to post answers to your own question but comments or edit your question. Nitpicking: see http://grammarist.com/usage/bound-bounded/

[abbax]

I'm sorry =)

unluckily I'm a novice in this topic so I don't know what to do and where to search.

For my purpose the only thing I've done is find the structures which have at least one ligand (here) but I would like to delete the structures having exactly one ligand.

PS: sorry for the language mistake

Answer 1

[David Westergaard]

Hi there!

So, first off, your question is a bit ambigous. "In particular I would like to search, on the PDB, the structures in which proteins are bound to more than one ligand" - this can be understood in two ways:

1. Actual crystal structures in the Protein Data Bank where more than one ligand is bound
2. Structures in the protein data bank for which there are multiple entries with different ligands

And actually, the answer to both of them is the same. There is, to my knowledge, no pre-computed database for this. What I would suggest is just downloading all the structures which has at least one ligand, and then just parsing them. You can read more about the PDB file format here. Specifically, in a first pass, you just have to skip anything which does not have a HET in the first column. Mind you, in some cases HET can be both peptides and small molecules.

You'll quickly notice that some ligands are not really interesting (phosphates, stabilizing agents etc, etc). From my experience, working with a lot of ligands and crystal structures generally requires a lot of manual curation.

Hope it helps!

Answer 2

[Jiyao Wang]

You can search in NCBI structure database: https://www.ncbi.nlm.nih.gov/structure/?term=2:300[BioUnit Chemical Count]