Thread by @theAliceRoberts

My friend Pontus Skoglund and his colleagues have JUST published this incredible new piece of research into Viking-era migrations - using ancient DNA and a brand new method called TWIGSTATS.

This epic piece of research is matched by an epic cover of first 2025 issue of Nature!

The cover art - by Oliver Uberti - is inspired by intertwined serpentine knots on Viking runestones and includes runes for the ‘letters’ of DNA: A, T, G, and C.

Additional runes for “man-journey-heritage” appear in the lower-right corner.

This new analysis includes a new piece of information relating to one of the stories in my latest book, Crypt: the remains of 30+ individuals from a mass grave in Oxford, thought to have been slaughtered in the 1002 St Brice’s Day Massacre, when Aethelred II used hate speech to incite ethnic violence against ‘Danes’ settled in England.

Ancient DNA analysis is an incredibly powerful tool for understanding populations, migrations and families in the past - archaeogeneticists are now in the business of reading entire ancient genomes. And this new Twigstats methods has made it even more powerful.

This is really exciting science. And it’s really hard. First the archaeogeneticists have to take samples from ancient bones, then they extract DNA from the samples, being very careful not to contaminate it with any other DNA.

They can boost the amount of DNA, getting it to copy itself many times until they have enough to work with. (They do this using a technique which also ends up as shorthand for DNA tests, like the one we got familiar with during the early years of the COVID pandemic: PCR).

Then the task comes of sequencing or ‘reading’ the DNA - deciphering the code, letter by letter. Ancient DNA might survive the ravages of time - but it’s usually broken up into tiny pieces, some just a hundred letters long - while a whole human genome is nearly 3000,000,000 letters long!

So the next stage is assembling those segments of code into a whole genome. It’s a mindblowingly huge puzzle which has only been made possible with advances in computing power.

But reconstructing an ancient genome is just the beginning. What the archaeogeneticists are really interested in is comparing lots of genomes - looking for similarities and differences. These are the key to understanding what was happening to populations in the past - and tracking migrations.

Over time, genomes acquire new mutations. Those mutations will be inherited by descendants of the person in whom the mutation first arose. While most of the genome will be identical between two people, the pattern of these little differences is crucial for working out what happened to populations over time - and for tracking families.

Archaeogeneticists have been able to compare genomes and work out how migration contributed to the spread of farming in the Neolithic, between 11000 and 5000 years ago, for instance. In the Stone Age, the differences between populations were relatively large and easy to spot.

But once you get to around 2000 years ago, into the Iron Age, there’s much less difference between populations. Pontus told me that distinguishing different Iron Age Scandinavian groups, for instance, is MUCH harder. It’s not a perfect analogy, but imagine trying to tell if someone is Norwegian or Danish by the appearance - it’s hard.

But Twigstats works by focusing on salient differences - and these are ones related to relatively recent mutations (with respect to the date of the individual sampled). It ignores older mutations which have been sifted and sorted through populations over time, creating confusing ‘noise’.

So it’s a bit like snipping a twig off a gene tree so you can focus on just those branches - and ignoring the rest of the tree. Hence the name: Twigstats.

Pontus and his colleagues tested their new technique on simulated data to make sure it worked - which it did, very well. Ten times better than anything else they’d tried before!

So they used it to look at some real samples. They looked at modern human and Neanderthal DNA - to test the theory that these species interbred with each other in the Palaeolithic.

Now this is something I’ve been following for years, ever since my first big landmark series for the BBC, The Incredible Human Journey. Back then, in 2008, there didn’t seem to be any evidence of modern humans (Homo sapiens) and Neanderthals mixing. At least, fossil bones weren’t providing a suggestion of interbreeding...

But then ancient DNA came along - with the first Neanderthal genome published in 2010. And then it seemed there WAS evidence for interbreeding, around 60,000 years ago.

But a few scientists have criticised this claim, saying that the patterns in the modern human DNA could have happened another way, just through genetic variants being sorted over time, creating a sort of mirage of an interbreeding event with Neanderthals.

Most archaeogeneticists still thought that interbreeding explained the pattern. But there was an element of doubt.

No longer! Twigstats finally puts that doubt to bed. Neanderthals and modern humans definitely met up and swapped genes with each other (delicately put). (painting by Tom Bjorklund)

Another finding published in the new paper - which mostly focused on Viking Age genomes, hence the beautiful cover of Nature - was that a 2nd-4th CE individual from the military/gladiator cemetery of Driffield Terrace in York, already known to have unusual ancestry, is now shown to have Scandinavian ancestry. That’s interesting because it shows how people were moving around quite a bit in northwest Europe in Roman times.

But the main focus of the study was the later first millennium - the Viking Age. And here, Pontus and his colleagues were able to use Twigstats to track migrations of different populations that have been difficult to ‘see’ before - because they’re all so similar.

They found evidence for migrations of Germanic people southward into Poland and Slovakia, as well as into south-central Europe and Britain.

They also found evidence of a previously unknown migration into Scandinavia, before the Viking Age - transforming the ancestry of Denmark and southern Sweden. Interestingly, this seems to coincide with a change in runic script and language - establishing Old Norse.

And finally, most of the individuals from the Viking Age mass grave in Oxford, that I wrote about in Crypt, were shown to have Scandinavian ancestries. It seems that they probably were ‘Danes’ killed on St Brice’s Day in 1002.

With a paper like this - announcing some new results and fascinating insights into medieval history - unveiled using a brilliant new technique, you can be sure that there will be more revelations to come!

Here’s a link to the paper:
nature.com/articles/s4158…