I’m guessing everyone reading this post is familiar with recent research from Svante Paabo’s group indicating that modern humans interbred with Neanderthals during their long co-existence in Eurasia between 30,000 and 80,000 years ago. According to the researchers’ calculations, somewhere between 1 and 4% of the DNA in modern non-African humans is derived from these interbreeding events – in other words, many of us are walking around with Neanderthal DNA sitting in our genomes.
So how much of your genome is Neanderthal? Over at The Genetic Genealogist, Blaine Bettinger takes a look at the options currently available to those interested in digging for Neanderthal ancestry in their own genetic backyard. Blaine notes that one company is already offering a test labelled as looking for Neanderthal ancestry based on a limited number of variable (microsatellite) markers. However, this test doesn’t actually look directly for putative Neanderthal-derived variants; instead, it (rather quaintly) tests for “strong matches between your DNA fingerprint […] and populations identified as “archaic,” that is, whose composition retains the earliest earmarks of out‐of‐Africa genetics.” This is a very rough approach to the problem, to put it mildly.
Added in edit 15/07/10: John Hawks has a justifiably scathing review of the test on his blog; I’ve removed links to the company from this post to avoid giving them extra publicity.
People who have already had their genomes scanned by a company like 23andMe theoretically have sufficient data already available to perform a much higher-resolution analysis. However, sadly there’s not yet any readily available algorithm out there for doing this, despite there being (as Blaine notes) substantial interest for such a test from amongst the 23andMe community.
Seems like there’s some real scope for a DIY genomics tool here. Is anyone out there already working on this? Let us know in the comments.
RSS
Email
Twitter
Authors Elsewhere
As far as I know, Paablo’s group haven’t released a list of variants that are believed the derive from Neandertals, so someone would have to ask them for that. Also, if I read the analysis correctly (which isn’t hugely likely, tbh, it was a very long supp.) I think they did something more complicated than just produce a list of such variants and look at them as a proportion of the genome.
So, really, what we’d need is a list of variants, each with the supposed Neanderthal derived allele and a posterior of it actually being so. Once you had that, though, the rest should be pretty easily. Someone will need to get into contact with them to get hold of it.
there are variants which they gave that they think are in high frequency in eurasians which are from neandertals. in the 23andme forums people use those SNPs to figure out if they’re neandertal or not at that locus.
Speaking as your friendly neighborhood bureaucrats, we wouldn’t advise releasing any “DIY” genome software without pursuing a multimillion dollar 510k application.
This message brought to you by the guardians of public health. What if someone threw themselves off a bridge because of one of these unregulated tests for Neanderthal ancestry? Please think of the Neanderthals.
Please don’t bother telling the Agency that genetic tests could be “just for fun” or that ancestry is not medically relevant. Of course it is.
I’d happily work on a comparative tool but I agree with Luke; I don’t believe that the data are available in a suitable form. I have waded through the 175-page “1.pdf” (please!) supplementary data, but there is no sign of the so-called “Neanderthal SNPs” discussed in the 23andMe forum.
Hi guys,
I can’t find the SNP list either in the Supplementary Data. However, there is one less direct approach.
You can download the list of SNPs that carry the derived (non-ancestral) allele in Neanderthals from UCSC (here’s the track schema), and here’s the direct link to the file (warning: big gzipped file). In the paper, the authors used these SNPs to show that non-African humans carried more derived Neanderthal alleles than Africans, thus indicating gene flow; in theory you could do the same calculation for a set of 23andMe data-sets to see who was more Neanderthalised. It’s a crude instrument, and doesn’t (as far as I can see) lend itself to an actual estimate of the % of Neanderthal admixture in your genome, but it’s a start.
There’s also a set of candidate regions for gene flow (again, gzipped file) that might be useful.
It will take some more digging in that Supp Data file to figure out the best way of doing this.
I’ve already gone to the trouble of tracking down the specific 23andMe SNPs that have a Neanderthal origin largely by combing through UCSC and other sources as Daniel described. This blog list all the relevant SNPs.
David Lonard
http://spittoon.23andme.com/2010/05/06/new-evidence-suggests-interbreeding-between-humans-and-neanderthals/
If you could assist me I would appreciate it. I have read lately of a lab looking for test subject with potential ‘ancient’ genes. The media was enjoying how many women submitted their husbands for a ‘cave man test’ if that helps narrow down the solicitation article.
I am unable to locate it and actually wish to contact them seriously about this issue. I actually DO have an unusual genetic makeup. The pediatrician told my adoptive parents I was a ‘throw back’ I believe it is properly known as atavistic gene expression? Suffice to say the ‘issues’ I had in my youth continue to manifest in unusual ways as an adult and many friends say I am an argument for some form of divergent evolution. Neandertal, Denisovan or whatever I would like to get to the root of this.