You can also use PICR (http://www.ebi.ac.uk/Tools/picr/), it has access to mapping tables for legacy identifiers such as IPI, but it is kept much more up to date than DAVID, which hasn't updated its files since 2010 or so.

It is pretty straightforward to use, too. I recommend limiting your serach by species to get a cleaner output and a faster query.

Go to http://www.uniprot.org/ and choose the "ID Mapping tab", then use the scroll down menu "From" IPI, "To" UniPrtotKB AC. For instance:

From To

IPI00003269 Q562R1

IPI00552546 Q9BTE3

IPI00792032 Q6TBL4

IPI00792032 E7EWT8

Thank you Javier, but the uniprot ID Mapping service don't allow the variations at the end of the IPI code (.1; .2; .3; etc). So, for some IPI codes you could get several UniprotKB AC codes (like in your example for IPI00792032).

DAVID has an gene ID converter which can convert IPI to uniprot

http://david.abcc.ncifcrf.gov/conversion.jsp

Using ID Mapping is, probably, the best option as it can be reproducibly used by other researchers which use the same data set. Mapping IPI to several UniprotKB codes is a result of protein redundancy rather than wrong mapping/annotations so it is better to take them into account. Regarding the .1.2.3 code you can remove the .number because it shows revision history rather than novel IPI id and then use ID Mapping.

Thank you for your contributions.

I understand the redundancy problem but with the Uniprot ID Maping there are hundreds of IPI codes without a corresponding Uniprot AC code!

Regarding the mapping tool of DAVID, I'm still not able to use it.

Just ran into this problem.  The David conversion worked for me to generate ENSEMBL IDs

Some IPI sequences do not appear in UniProt for good reasons. As IPI is discontinued since 2011 you should not use it. Depending on the species you can lose up to 20% of the sequences in IPI, although those are unlikely to have existed in reality. ID mapping on the uniprot website is no longer possible and the mapping are no longer maintained. 

http://www.uniprot.org/news/2014/01/22/release

Article Consequences of the discontinuation of the International Pro...

Jerven, that does not sound like good news to me, but thanks for your important note.

It is important for anyone who asks this question to know that IPI curation has ceased years ago, that all of the old converters that I have tried no longer work, and that UniProtKD IDs must be used instead of IPI IDs.

Hi Nuno,

I used Benjamin suggestion and it worked; thank you Benjamin.  I was able to convert the IPI accession numbers to UNIPROT_ID  by selecting for ENSEMBLE_GENE ID  under "Select Identifier" pulldown menu, at David Bioinformatics site, using Gene ID conversion Tool.

Best, Hediye.

You can also use PICR (http://www.ebi.ac.uk/Tools/picr/), it has access to mapping tables for legacy identifiers such as IPI, but it is kept much more up to date than DAVID, which hasn't updated its files since 2010 or so.

It is pretty straightforward to use, too. I recommend limiting your serach by species to get a cleaner output and a faster query.

Hi Pablo,

Could you please help in identifying the latest update of PICR, please?

I also converted IPI in Uniprot IDs (today), but it would be important to know the release of both databases that PICR used for this conversion.

Many thanks in advance!

Hey Pablo & Annarita, seems that I'm dumb! I could not use the PICR software! Can you please guide?? 

Many thanks.