PDB to Uniprot Mapping

Question

[stackf05]

Hi

I have downloaded the pdb_chain file from sifts.

However, I note that one UniProt entry can have multiple pdb code. Yes it is possible to do the mapping. But how the 3D structure of a protein is determined for that specific UniProt entry when it has multiple pdb_code?

ftp://ftp.ebi.ac.uk/pub/databases/msd/sifts/text/pdb_chain_uniprot.lst

Help please. Thanks.

Answer 1

[Jean-Karim Heriche]

I can think of several reasons why a UniProt entry is linked to multiple PDB ones. First, the data for the protein could come from multiple studies. Second the protein could be linked to different structures corresponding, for example, to mutations or different complexes.

[stackf05]

But which 3D structures should I display then?

[Pallab Bhowmick]

That depends on your problem. But I will prefer unbound state PDB structure (best structure among multiple unbound PDB) for Uniprot entry!!

[stackf05]

Now what's unbound state PDB ?

[Elisabeth Gasteiger]

Just to confirm the above reply:

Indeed, a PDB identifier matches a particular experiment. It happens frequently that several structures are determined for a particular protein in UniProtKB: wild-type vs mutant, apoprotein vs structure of protein bound to ligand A, etc

see also http://www.uniprot.org/help/multiple_pdb_xrefs