Which Are The Most Frequent Ligands In The Pdb Database?

Question

[Flow]

I mean, the ligands that bind with most different proteins. How could I get this information? It belongs to a previos question, where there are too many questions already, so I prefere to ask it separatedly. Thanks

Answer 1

[Michael Kuhn]

PDBsum has a mapping from ligand codes to PDB files (under "Ligands"), so with a bit of counting you have what you need.

[Michael Kuhn]

PDBsum is quite a useful resource, so I figured they might also have this kind of information

[Flow]

yes, this is what I was looking for! thanks, BTW how did you find it?

Answer 2

[Lyco]

As usual, a note of caution from me: Don't forget that whatever census you use (PDBsum is indeed a good start), the number will be heavily biased by a sampling artefact. Most likely, you will not get 'the ligand that binds to most different proteins' but rather the ligand that binds to the proteins that have the best coverage in your database of choice. If you use PDB-derived data, contributing factors include i) how suitable are the proteins for structural studies (crystallize well? small enough for NMR?) ii) the importance (how many independe entries for the same protein), iii) phyletic range (same protein from hundred of different species) and lots of others. Then there are the 'artificial ligands' (i.e. stuff people put into their crystallization buffer) and the boring ligands (Zn, Water, ATP). These things should be kept in mind.

Apart from that , I second the idea to use PDBsum. Another recource is iPFAM (which is currently hibernating but can be downloaded by ftp)

Answer 3

[Khader Shameer]

Have you tried LigBase or PDBeChem ? You can also get the ligand data in PDB format from PubChem or ChEBI.