Identifying which allele is derived in non-human species

Question

[Rubal]

Hello Everyone,

I have a sets of SNPs from multiple non-human species and I would like to know for each one which of the twos allele is ancestral and derived. I remember doing before and I think on UCSC genome browser website there was a way to visualise for a given site what that allele was in several other species, so one could infer which allele is likely to be ancestral. However currently I can't find a way to do this on the Ensembl or UCSC website.

If anyone can recommend a way to find out the ancestral allele for a given site, either in an automated way for a batch of SNPs or for single sites in a web browser that would be very helpful.

Similarly, if there is a straightforward way to get the conservation score from one of the many conservation tracks that would be helpful. I know for human data I have used wiggle tracks for conservation, which I found tricky to handle and I wasn't confident that it was working correctly. Is there a good source for finding the conservation score for a batch of SNPs? Ensembl's Biomart doesn't seem to offer this feature, or I missed it.

Help for either or these issues (derived/ancestral allele or conservation score for batches of SNPs) would be greatly appreciated.

Thanks in advance!

Answer 1

[Emily]

If you search for a variant ID in Ensembl, you should find the ancestral allele near the top of the page. You will also find the base in other species if you go to the phylogenetic context page.

Ancestral allele is a available as an Attribute in the Short Variation database in BioMart. Just filter by your list of IDs, for example in this query.

[Rubal]

Thanks so much, wasn't aware of this, I'll check it out. Is this also possible for arbitrary genomic positions eg chr1:100 without a variant ID? My data comes from VCFs where I have genomic coordinates but not variant IDs, I'm not sure how to know if any of my coordinates correspond to known variant IDs. But it looks like I can achieve a lot of what I want with the phylogenetic context page. Thanks very much.

[Rubal]

Is there a way to get ancestral allele information for a site that is not a known SNP? I have some sites that are polymorphic in my population but not known variants in the database and I'd like to know the predicted ancestral allele. For an individual site I can see the eutharian mammals 39 way alignment for a site in the browser but I'd like a way to download this or simply an inferred ancestral allele for a list of positions. Thanks very much for your help.

[Emily]

There's a standalone script here that does that.

[Rubal]

Thank you very much for pointing this out