[Biodevelopers] Mapping between various protein indices

Dan Bolser dmb at mrc-dunn.cam.ac.uk
Tue May 25 09:26:23 EDT 2004

On Tue, 25 May 2004, Joke Reumers wrote:

>Maybe a bit out of scope on this mailing list, but I'm trying to find a
>complete mapping between NCBI RefSeq (of Genbank gi) indices and
>SwissProt/TrEMBL accessions. The best I could find so far is the IPI index,
>but this seems far from complete. Anyone any better ideas?

Hi, I had this exact same problem recently! The 'best' two solutions I
finally found were;

1) Use SRS to map from RefSeq -> uniPark -> SwissProt

(can't find where to get unipark from directly)

2) Use SeqHound to get SwissProt from GI.

I was shown the SRS method, e.g.


And I got the following instructions on the use of SeqHound...

If you want a list of GI's that have *exactly* the same sequence as your
input protein GI, then you could use

The Perl version of this takes a GI as input and then returns a list of
GI's.  You could use SHoundSeqIDFromGi (or SHoundSeqIDFromGiList) to find
the Swiss-Prot sequence.

Using either of these should get you a full mapping. Let me know how you
get on, 


>Thanks in advance,
>Joke Reumers
>Biodevelopers mailing list
>Biodevelopers at bioinformatics.org

More information about the Biodevelopers mailing list