[ensembl-dev] Orthology information

[Matthieu Muffato]

Dear Ogan,

I think it is an error in our pipeline to miss the orthology link 
between Ccr5 to CCR5.
With our current pipelines, we are unfortunately unable to force given 
genes to be orthologous, which would fix the issue. We are working on 
having this feature, but it won't be available before next year.
In the meantime, I am afraid such discrepancies in the data cannot be 
avoided

If you look at the whole tree, among the 22 human genes: CXCR1 and CXCR2 
on one side, and CCR2 and CCR5 on the other side are mixed up. 
Otherwise, all the human and mouse genes are grouped together correctly.
http://e77.ensembl.org/Homo_sapiens/Gene/Compara_Tree?collapse=12110668%2C12110661%2C12111294%2C12111302%2C12111538%2C12111288%2C12110772%2C12111571%2C12111144%2C12111233%2C12110861%2C12111503%2C12111610%2C12111426%2C12111097%2C12110887%2C12111121%2C12110805%2C12110994%2C12111194%2C12111071%2C12111209%2C12111103%2C12111377%2C12110798%2C12111300%2C12110704%2C12110826%2C12110620%2C12111562%2C12111433%2C12111234%2C12111641%2C12111006%2C12111176%2C12111081%2C12111198%2C12111643%2C12111576%2C12111470%2C12110998%2C12111375%2C12110687%2C12111340%2C12110880%2C12111389%2C12111279%2C12111257%2C12111677%2C12110765%2C12111077%2C12111436%2C12111100%2C12111193%2C12111152%2C12111585%2C12110758%2C12111455%2C12110697%2C12111409%2C12111131%2C12111620%2C12111108%2C12111670%2C12111147%2C12110948%2C12111687%2C12110816%2C12111125%2C12111134%2C12110941%2C12111376%2C12111085%2C12111295%2C12110886%2C12110791%2C12111229%2C12111551%2C12111397%2C12111559%2C12111067%2C12111588%2C12111141%2C12111627%2C121
 11008%2C1
2110782%2C12110952%2C12110756%2C12110742%2C12110964%2C12111241%2C12111055%2C12111646%2C12111039%2C12110664%2C12111289%2C12111526%2C12110642%2C12111464%2C12111675%2C12111162%2C12110991%2C12111284%2C12111200%2C12110761%2C12111253%2C12110932%2C12110673%2C12110677%2C12111196%2C12111479%2C12110654%2C12110817%2C12111136%2C12110827%2C12111684%2C12110890%2C12110971%2C12110781%2C12111363%2C12111692%2C12110692%2C12111534%2C12110726%2C12110838%2C12111111%2C12111224%2C12111441%2C12111664%2C12111567%2C12111313%2C12111161%2C12111510%2C12111505%2C12111374%2C12111010%2C12110731%2C12111554%2C12111420%2C12110684%2C12111274%2C12111385%2C12111174%2C12111404%2C12111243%2C12110869%2C12110873%2C12111353%2C12110909%2C12111263%2C12111655%2C12111536%2C12111393%2C12111207%2C12110901%2C12111026%2C12111251%2C12110656%2C12111462%2C12110927;db=core;g=ENSG00000121807;g1=ENSMUSG00000079227;r=3:46353734-46360928

In both cases, the order of the duplications at the root of mammals is 
not clear, and as a result, the definition of the sub-families is 
partially wrong for the human and mouse members (it is correct for other 
mammals, though)

Sorry for the inconvenience,
Matthieu

On 20/10/14 04:13, Ogan Mancarci wrote:
> Hi,
>
> I was using the orthology information in Ensembl for my research when I
> noticed there were some mismatches like mouse Ccr5 matching to human
> CCR2 instead of CCR5. Is this a known issue? From other sources I can
> match Ccr5 to CCR5 with high quality.
>
> Cheers,
> Ogan Mancarci