Which Downloadable Database Of Pdb Ligand Binding Sites Would You Recommend?

Question

[Jan Kosinski]

I am searching for a database of protein ligand binding sites in PDB that

• is download-able
• is up to date with PDB (not older that approx. one month)
• lists residues interacting with that ligand

If I could get interaction types like in PDBe (e.g. http://www.ebi.ac.uk/pdbe-site/pdbemotif/?tab=boundmolecule&pdb=3mht&ligandCode3letter=SAH) I would be delighted.

Is anything out there that you could recommend?

[Nir London]

Can't you perform an 'advanced' query and batch download the results?

[Michael]

Is this not in PDBs FTP site? http://www.pdb.org/pdb/static.do?p=download/ftp/index.html

[Jan Kosinski]

I think ligand binding site information is not there. I think the binding site data from PDBe is in the PDBe Motif, but I am not sure yet how to download and handle that.

[Jan Kosinski]

Nir, I think I should be able to that using the PDBeMotif web interface, but I simply have no idea how. I queried EBI HELP for that, but I am still waiting for a response.

[Jan Kosinski]

Got the response from EBI HELP. Local installation of PDBeMotif is too difficult and too hardware demanding. And the newest software won't be available. I got some hints about XML queries of PDBeMotif online, but I am not sure yet if it is possible to get all data I need.

Answer 1

[Joerg Kurt Wegner]

The CREDO database might be a solution, and you might need to check-in with Adrian Schreyer for the update cycles.

[Jan Kosinski]

CREDO was may favorite, as it has nice python interface, and can be installed easily. I queried Adrian Schreyer a couple of days ago. He says that he will soon do the update (the current version is from July 2010).

[Joerg Kurt Wegner]

Cool, that is what I was hoping for!

Answer 2

[Tommy Carstensen]

Can you use this one? http://www.bigre.ulb.ac.be/Users/benoit/LigASite/

[Jan Kosinski]

Thanks, I will try it out. It seems to be updated regularly. The downside is that they use some filtering procedure to filter out small, low contacting ligands (http://www.bigre.ulb.ac.be/Users/benoit/LigASite/index.php?methods). I will look at its content in more detail and post the results here.

[Jan Kosinski]

I have looked at LigASite more carefully and it does not serve my purposes. First, it includes ligand sites only for those proteins for which there is at least one apo- and one holo-structure. Second, it misses some data even if apparently there is both apo- and holo-structure available (e.g. 3kdk).

Answer 3

[Khader Shameer]

Have you tried Catalytic Site Atlas (curated data) ?

[Jan Kosinski]

To my knowledge, Catalytic Site Atlas is just for catalytic sites, not ligand-binding sites.

[Khader Shameer]

IMHO, CSA fits the flow's requirement (download-able, is up to date with PDB (not older that approx. one month, lists residues interacting with that ligand).

[Jan Kosinski]

I downloaded the newest CSA dat file from here and the dat file includes only catalytic residues. I am guessing that the only ligand-interacting residues that are listed there will be those that have also been annotated as catalytic.

Answer 4

[Vladimir Chupakhin]

Try sc-PDB - binding site and proteins are already prepared for docking.

Info from the site: To assist structure-based approaches in drug design, we have processed the PDB to identify binding sites suitable for the docking of a drug-like ligand and we have so created a database called sc-PDB. The sc-PDB database provides separated MOL2 files for the ligand, its binding site and the corresponding protein chain(s). Ions and cofactors at the vicinity of the ligand are included in the protein.

[Jan Kosinski]

Unfortunately, on the website they say that "The database is updated annually", I need sth much up to date.

[Vladimir Chupakhin]

Yep, so at this point is only way - is to prepare it by yourself (also take a look at the preparation of the sc-PDB it might help you).

Answer 5

[Jan Kosinski]

After getting help from EBI support I got a db of ligand binding sites from PDBeMotif database working, quite up-to-date and local.

I am using it for some time already, and it seems to be cool. It basically lists all bonds to hetatms from PDB, together with distance info, and with annotations on interaction type.

I wrote some quick-and-dirty Python code to download and use this db.

Installation

pip install http://modorama.biocomputing.it/beta_static/src/pdbemotif-0.1.0.tar.gz

It should install pdbemotif Python module, and two scripts should go to your python bin:

download_selected_lig_sites.py
download_all_lig_sites.py

They download data on bonds with protein and hetatms. You can use like this:

python -u path_to_bin/download_all_lig_sites.py /data1/outdir/
python -u path_to_bin/download_selected_lig_sites.py some_pdbcodes.txt /data1/tmp/outdir/

where some some_pdbcodes.txt file lists pdbcodes, one per each line.

Downloading all "sites" takes 3 days. You will get a list of xml files, one per each pdbcode.

Then you can do:

from pdbemotif import BoundMoleculeSite
site_file = open('/data/db/PDBeMotif/3mht.txt')
mol_site = BoundMoleculeSite(site_file.read())
site_file.close()
print mol_site.get_all_binding_resi_for_chain('A')

[Khader Shameer]

He already mentioned about PDBeMotif in his question. AFAIK, he is looking for similar/related resources.

[Khader Shameer]

I think @flow already mentioned about PDBeMotif in his question. AFAIK, he is looking for similar/related resources.

[ajingnk]

Could I still borrow your code? I try the command you pose, but it gave ERROR 404: Not Found. Thanks!

Answer 6

[Adrian]

The new version of the CREDO database has downloadable data sets of interatomic interactions/structural interaction fingerprints. Currently, there is a comprehensive kinase set but I could create others if there is enough interest.

[Dan]

Sounds interesting to me :-)