Map The Protein Sequences Of Unknown Genes That Have Uniprot Names To Ensembl Genes.

Question

[Hmm]

Is there a way to map the protein sequences of unknown genes that have uniprot names to ensembl genes. (command line or through api or any ways....)

Uniprot names are as follows:

YJ015_HUMAN
ZNAS2_HUMAN
CJ052_HUMAN
YJ004_HUMAN
CEAS1_HUMAN
YJ008_HUMAN
CTLFB_HUMAN
YJ018_HUMAN
ZN487_HUMAN
FRP2L_HUMAN
SRCRM_HUMAN
HUG1_HUMAN
TI23B_HUMAN
ADAS1_HUMAN
YJ001_HUMAN
DMBTL_HUMAN
YJ017_HUMAN
YJ016_HUMAN
CSC2A_HUMAN
ASA2C_HUMAN
ANTRL_HUMAN
AGA10_HUMAN
CJ085_HUMAN
CJ136_HUMAN
CJ108_HUMAN
CJ040_HUMAN
F25DE_HUMAN
YJ012_HUMAN
CJ115_HUMAN
YJ006_HUMAN
DAS1_HUMAN
CJ112_HUMAN
AKCL1_HUMAN
FA22B_HUMAN
CJ126_HUMAN
YJ013_HUMAN
CCYL2_HUMAN

Answer 1

[aravind ramesh]

Hope this helps You. UniProt ID Mapping. Download the text file from this site and use it As you like(via a program). You can also get data by Organism wise too.

[Hamish]

The UniProt.org Identifier Mapping service also provide a REST Web Service, see http://www.uniprot.org/faq/28#id_mapping_examples for details.

Equivalent functionality is also available via the EMBL-EBI's Protein Identifier Cross-Reference (PICR) service which provides both REST and SOAP based Web Services for automation.

Answer 2

[Andy Yates]

Hi,

As the person who works very closely with Ensembl's xref system. Ensembl's UniProt xrefs come from two sources. The first is direct association where a UniProt annotator has made an assertion about the link between us & them. The second is an alignment using Exonerate. I haven't investigated your list in that much detail but the random selection I took had the following results

• CJ108_HUMAN redirects to http://www.uniprot.org/uniprot/Q8N8Z3. Seems like a retired identifier. Q8N8Z3 does link to Ensembl (ENSG00000180525)
• YJ015_HUMAN has a note saying Product of a dubious CDS prediction. Protein evidence level 5 (uncertain - not good)
• CJ052_HUMAN has a note saying Product of a dubious CDS prediction. Probable non-coding RNA. Protein evidence level 5 (uncertain - not good)

I would recommend pre-filtering this list through UniProt a trying to reduce it down to a set of proteins which you think are really missing from Ensembl. Filtering those which are retired and maybe consider if you really want to look for PE level 5 proteins. Then the best action I can suggest is to use Ensembl's Blast service to find the mappings. If you have any problems please contact our helpdesk and I'm sure we can help out a little more.

[Hamish]

UniProtKB names are subject to change, in the case of CJ108_HUMAN the name has changed to PRR26_HUMAN, but the primary accession (Q8N8Z3) has remained the same (see http://www.uniprot.org/uniprot/Q8N8Z3?version=* for a summary of the changes to this entry). Unsurprisingly UniProt recommend the use of accession numbers over names, although names are slightly more human friendly.