[ensembl-dev] Question about missing regions in whole genome alignments

[Matthieu Muffato]

Dear Marc,

Thank you for bringing this to our attention. I have looked in some of 
the debug output of the pipeline and I can see that Enredo thinks this 
human region is only found syntenic to cat in the 17-way EPO, so the 
error goes quite early in the pipeline. I still need to check whether 
the anchors were loaded correctly, but the overall coverage stats look 
fine. Unfortunately, we don't have all the files for the 8-way EPO. I'll 
let you know if I find out what happened to this region

Matthieu, Ensembl Compara

On 04/11/15 12:47, Marc P. Hoeppner wrote:
> Dear EnsEMBL developers,
>
> I've noticed that there could be a problem with certain whole genome
> alignments in the current (E82) version of EnsEMBL.
>
> Specifically, for the human region 17:58663933-58739040 (and possibly
> others).
>
> This region is syntenic on all primate genomes (8 primates EPO) as well
> as 23 amniotes (amniotes Pecan). However, in 17 mammals (EPO), this
> region is only identified as syntenic between human and cat - which I
> suspect is a mistake. I understand that Pecan and EPO use different
> algorithms, but the gene in this region (TEX14) is conserved across all
> mammals (and most vertebrates) and has no obvious homologs that may
> explain off-target alignments.
>
> Regards,
>
> Marc
>