search this blog

Friday, May 22, 2020

Genetic continuity and change in prehistoric Ireland (Lara Cassidy Thesis)

Over at TARA at this LINK. Here's the abstract:

This thesis provides an initial demographic scaffold for Irish prehistory based on the palaeogenomic analysis of 93 ancient individuals from all major periods of the island's human occupation, sequenced to a median of 1X coverage. ADMIXTURE and principal component analysis identify three ancestrally distinct Irish populations, whose inhabitation of the island corresponds closely to the Mesolithic, Neolithic and Chalcolithic/Early Bronze Age eras, with large scale migration to the island implied during the transitionary periods. Haplotypic-based sharing methods and Y chromosome analysis demonstrate strong continuity between the Early Bronze Age and modern Irish populations, suggesting no substantial population replacement has occurred on the island since this point in time. The Mesolithic population shares high genetic drift with contemporaries from France and Luxembourg and shows evidence of a severe inbreeding bottleneck, apparent through runs of homozygosity (ROH). Substantial contributions from both Mediterranean farming groups and northwestern hunter-gatherers are evident in the Neolithic Irish population. Moreover, evidence for local Mesolithic survival and introgression in southwestern Ireland, long after the commencement of the Neolithic, is also implied in haplotypic-analysis. Societal complexity during the Neolithic is suggested in patterns of Y chromosome and autosomal structure, while the identification of a highly inbred individual through ROH analysis, retrieved from an elite burial context, strongly suggests the elaboration and expansion of megalithic monuments over the course of the Neolithic was accompanied in some regions by dynastic hierarchies. Haplotypic affinities and distributions of steppe-related introgression among samples suggest a potentially bimodal introduction of Beaker culture to the island from both Atlantic and Northern European sources, with southwestern individuals showing inflated levels of Neolithic ancestry relative to individualised burials from the north and east. Signals of genetic continuity and change after this initial establishment of the Irish population are also explored, with haplotypic diversification evident between both the Bronze Age and Iron Age, and the Iron Age and present day. Across these intervals selection pressures related to nutrition appear to have acted, with variants involved in lactase persistence and skin depigmentation showing steady increases in frequency through time.

CASSIDY, LARA, A Genomic Compendium of an Island: Documenting Continuity and Change across Irish Human Prehistory, Trinity College Dublin.School of Genetics & Microbiology. GENETICS, 2018

Update 24/05/2020: Apparently the thesis has been embargoed until 24/5/2023. It's still marked as open access, but the link to the PDF now leads to a login page.

See also...

Commoner or elite?


Romulus the I2a L233+ Proto Balto-Slav, layer of Corded Ware Women said...

highly inbred
They've found an ancient Habsburg! Charles the 2nd of the Neolithic.

Samuel Andrews said...


All of modern "Atlantic" Europe mostly derive from Bell Beaker. But, we have all known for a long time that Ireland & Basque have the highest amount of local continuation going back to Bell Beaker/Early Bronze age.

But, there was for sure a shift in Ireland in Iron age due to Celtic invasion (who themselves also came from Bell Beaker). I think amateur experts will come close to answering what change happened.

Bob Floy said...

So, a trickle from Hallstatt(indirectly), and a sprinkle of Viking, apart from that modern Irish are essentially Rathlin island. Love it.

AWood said...

Despite not being engaged member in the methodologies of the field, the thesis reads very well, almost like a historical novel and is engaging. A lot of good information on ancient Ireland. I found the discovery of 2 Linkardstowns cists being H2a quite interesting and I immediately thought of the Druids, and wondered if these were elites who had retained a Middle Eastern identity amongst the swath of WHG I2a who gradually adopted agriculture. Later on it is discussed other Linkardstown cists are of the more common I2a type, but I would need to take a deeper dive to confirm. The steppe information isn't up to date, but that is to be expected in a rapidly changing field.

Samuel Andrews said...

Page 196 shows phenotype allele frequencies in Mesolithic, Neolithic, Bronze, iron age Ireland.

Lactose persistent allele (rs4988235-A)
Bronze age=5%
Iron age=50%

Light skin allele (rs16891982-G)
Bronze age=70%
Iron age=100%

Steven said...

The study says that the Caucasus related component is present in Yamnaya related populations at a proportion of 75-80% of their total I incorrect to think this is wrong?

Davidski said...


Yep, 75-80% sounds like nonsense. This is probably based on an ADMIXTURE run.

Samuel Andrews said...

I wish they tested for red hair SNPs. Because, modern Northwest Europe has around 10x higher frequency of red hair alleles than Bell Beaker did believe it or not. Red hair is rare today, but used to be 10x more rare in the Bronze age.

So signature Irish red hair developed in British Isles, it must have happened between Bronze and Iron ages.

Romulus the I2a L233+ Proto Balto-Slav, layer of Corded Ware Women said...

The same general trends were observed, with HGs from across Europe estimated to be genetically taller than contemporaneous
Iranian, Anatolian and Georgian individuals. Clear variability in genomic height exists across European
HG groups, possibly with some geographical trends, although samples sizes are too low to ascertain
these reliably. Northwestern samples from the Rhine Basin (Bichon and Loschbour) show the highest
values, while Irish HGs display somewhat reduced scores, despite their close relation to these individuals.
Swedish, Spanish and Russian individuals display the lowest scores observed for European HGs, though
the risk scores estimated were still net positive.

That's really interesting that the Rhine Basin (Netherlands) and Balkans is where the genetically tallest HGs come from, and this is exactly where height in Europe peaks today i.e. among Netherlands and Western Balkans, Western Balkans obviously showing more HG continuity than in the East. Very obvious to anyone with a brain Europeans get their height from HG DNA and not the Steppe.

Rob said...

Awesome study with amazing transect.
Impressive attempts to link aDNA & social structure; e.g. differential lineage distribution of distinctive I2a by burial type is well spotted
Good discussion of linguistic issues & possibilities

Some minor side-points:

- ''it is perhaps noteworthy that none of the four Early Cardial Neolithic individuals so
far sampled (Croatia and Spain) belong to G2. Indeed, two possess haplogroups rare or absent in Balkans and Hungarian Neolithic (R and E1b1).''
-> R1b-V88 is common in S.E.E , E1b1 is simply not super common anywhere during the Neolithic, & IIRC one of the Tripoljes is E1b1

- Magdalenian had long ended by the Younger -Dryas phase

Samuel Andrews said...


Yeah, they do show all European hunter gatherers taller than Middle Eastern farmers. Measuring height from DNA is still tricky. Also, Loschbour & Cheddar man were really short, both around 5'4.

Romulus the I2a L233+ Proto Balto-Slav, layer of Corded Ware Women said...

When total length of the genome under ROH is considered, Newgrange10 shows a value close to the
average of sibling offspring (Fig. 3.9B). However, Newgrange10 has an inflated number of ROH tracts
compared to both sibling and half-sibling offspring and thus the average length of ROH is somewhat
diminished relative to sibling offspring. That said, the values seen for Newgrange10 are securely in the
range of that seen for sibling offspring and among the highest values achieved for half-sibling offspring.
No allowance for mutation or genotyping error was made during inbreeding simulations, which work to
break up longer tracts, and is the likely reason for the comparatively inflated numbers of ROH segments
seen in Newgrange10 and their diminished average length.

Guy boinked his sister.


Like I said, obvious to anyone with a brain

Samuel Andrews said...

It would make sense height in Europe comes from Mesolithic hunter gatherers. Then you how does one explain why Yugoslavia is tied for tallest in Europe but are average in hunter gatherer ancestry.

Yugoslavia and Baltic states aren't most developed economically but are very tall, so this has to be genetic height.

Italy, Sardinia, Spain under average height and highest in farmer ancestry. So, I agree there does seem to be a correlation.

Rob said...

Interesting about the east shifted Late iron age outliers

Matt said...

There are some really interesting bits here - I'm a bit disappointed that the haplotype data isn't more decisive, but such is the way of haplotype data. Highlights seem like:

- Lesser Magdalenian related ancestry in Irish HG, so probably out of refugium in Italy (total division of ancestry relating to the "Villabruna" clade between SE European and Italian refugium still pretty unclear to us though).

- Irish HG form a clade to exclusion of continental WHG, suggesting isolation, backed up by RoH.

- A surviving Neolithic ancestry rich hotspot in LCA-EBA SW Ireland. Suggests that Northern BB and/or following groups were not always exclusive in interactions with local EEF groups, though this is the edges of distribution. The slight excesses of outgroup f3 affinity by LCA-BA to the specifically Irish Neolithic set (relative to TRB and Iberia_LNBCA) is interesting as well, though could have multiple reasons (even including capture/shotgun differences).

Unfortunately there is still this gap in sequence around 2500 BCE (although this has more early 3rd m BCE than Olalde's paper), so I think this paper sheds not much light on whether integration of people from British+Irish neolithic is a bit higher than Reich and Olalde's group thinks (not very high, but maybe 10% higher, so more like 18-25% overall than 8-15%; mostly mainly women, though there is a I2 here that suggestive of some low level male incorporation).


I think this is all clearly written by Cassidy in 2017 *before* widely published Great Britain, etc. data, so hopefully the actual papers she publishes on these samples will be able to use samples like Mesolithic Great Britain HG to draw even more conclusions.

Targamos the Based, son of Kavkasos son of CHG son of said...

"Yugoslavia and Baltic states aren't most developed economically but are very tall, so this has to be genetic height."

Off the top of my head Albanians are pretty similar to Montenegrins and Serbs genetically yet are way shorter.

epoch said...


"Lesser Magdalenian related ancestry in Irish HG, so probably out of refugium in Italy"

My working hypothesis is that Magdalenian HGs took a massive hit from the Laachersee eruption. That would leave the fringes pretty much intact: Iberia, NW European cultures such as the Crosswellian and some groups in NE Europe.

There was a paper on the survival of considerate Magdalenian ancestry in Iberia. One of them was associated with the Azzilian and assigned a Y-DNA of pre-I1. I think NW European I1 was brought north by survival of small amount of Magdalenian ancestry there. The paper showed that neolithic groups showed small amounts of Magdalenian ancestry, Globalar Amphora being one of them.

DISCLAIMER: I have some bias here as I tend to really, really love Magdalenian art.

Ryan said...

I see my R1b-M222 ancestors are well represented. It shows we were established in NW Ireland pretty early FWIW.

Samuel Andrews said...


All encompassing, Supplemebtary Tables of this paper. Created by Anthorgenica user Pribislav.


They do have a lot of samples from circa 2000 BC. We have data from Beaker Netherlands and Britain which we can use as reference for first Beakers in Ireland. So, a model of Beaker Britain+Neolithic Ireland, should say how much local ancestry the Early bronze Ireland had.

The samples with high levels of Neolithic farmer ancestry in Southern Ireland is interesting. It would make most sense they have lots of local Neolithic ancestry instead of being from Spain.

Samuel Andrews said...

@Ryan, Yeah, your haplogroup must come from a Iron age clan who dominated that part of Ireland.

Someone at Anthrogencia posted a map showing 20-30% of people in Northern half of Ireland belong to R1b M222. Which is a lot more frequent than I would expect for a "Iron age" haplogroup.

I am sure with help with ancient DNA, people can understand the history of different R1b L21 clades in British Isles.

Samuel Andrews said...

As far as I know.....

70-80% of modern Irish belong to R1b L21.

Which means 20-30% of their Y DNA arrived after Early Bronze age. A deep assessment of modern Y DNA which isn't R1b L21, could get a good idea of where admix in "recent" history came from. I bet most of it is from England.

It seems every region of Europe, has Y DNA from every region of Europe. For example, E1b V13 in Scandinavia and I1 in Sicily and so on. All this, could be from historical intermarriage.

Btw, Vikings probably didn't contribute significantly to anyone outside of Iceland.

Romulus the I2a L233+ Proto Balto-Slav, layer of Corded Ware Women said...


Does the paper say that the Neolithic Y-DNA disappeared completely? These subclades of I2 make up 7.5% of the U.K. BioBank Y DNA. Definitely rare, but 1 out of 13 British Men is far from extinct. I expect Ireland originally paralleled the rest of the isles in that distribution but perhaps low frequency diversity was lost in the migration to North America.

Rob said...

@ Epoch
Lacher See eruption post dates Bichon , however, by 2000 years

Samuel Andrews said...


You have to breakdown what kind of I2 UK bank has. I suspect a lot of it isn't from Neolithic Britain. Some could be I2a2a1b2a-L801 which is Central European type brought by Anglo Saxons.

As we can see from, something like 50 Y DNA samples from Bronze age Britain & Ireland (hundreds of years after Bell Beaker), 0 are I2a. Yeah, there was basically 100% Y DNA replacement. We don't need modern Y DNA to see that.

Samuel Andrews said...


That's a good question to ask. But, Y DNA and mtDNA studies went out of style in 2010. Which is frustrating because the stuff that was published back then is crap.

Most of new info on modern Y DNA comes from FTDNA projects. And Only the people in those projects known anything about it.

What is I2a? We got to break down subclades.

There's not 7.8% though Neolithic Y DNA in UK. I can't believe there's that much Neolithic British Y DNA left. Because, it is absent in bronze age DNA.

Davidski said...


It looks like they've just embargoed the thesis until 2023. Even the abstract has been taken off.

The link to the PDF is still up there but it can't be accessed without a login.

epoch said...


Yes, true.

He also has no Magdalenian admixture and the site was at a mesh of lithic raw material exchanges if I understand correctly. Assigned to the Azilian culture, a proposed successor of the Magdalenian.

Yes, you are right, stuff can't be as straight forward as Magdalenians, boom, WHG. But still think the demise of the ancestry was connected to the Laachersee eruption, even if we have a half Magdalenian sample also assigned to the Azilian culture.

Samuel Andrews said...

@Davidski, That's annoying. We heard rumors of this paper 3 years ago. They finally make a thesis, then embargo it for three years. We might never get the DNA data.


Western European WHG is from Italian refugium. Eastern European WHG is from Balkans refugium. Azillian represents expansion of WHG out of Italy into mainland EUrope.

Sicily and Central Italy WHGs are basal to all WHG.

Ric Hern said...

So is it safe to say that Celtic formed during the Middle Bronze Age ?

Matt said...

A copy of this thesis can be found on this PDF hosting website,, search under Lara Cassidy, if anyone who has not still wishes to read it.

Kristiina said...

Regarding the survival of Neolithic I2 lineages in Ireland, this is what the paper itself argues: "Interestingly, the only Irish I2 lineage observed from the Beaker horizon onwards belongs to a Chalcolithic sample from the southwest of the island (Killuragh1). This individual belongs to the clade I2-M284, which is somewhat restricted to Britain and Ireland today and common in Irish Late Neolithic samples ... The two I2 lineages observed in the British Beaker and Bronze Age periods also belong to M284, supporting some small level of male line continuity across the transition on the islands".

Matt said...

@Kristiina, yeah, I saw that too. It's possible that is not continuity though, as such lineages may have survived among groups in Northern France or something like this, then re-entered Britain and Ireland via that route. It's tricky to work these very fine-scaled dynamics.

Kristiina said...

The conclusion made in the paper is more parsimonious. We have some ancient yDNA data from LBA and IA North France; Gurgy Les Noisats tumulus in Yonne, 300–100 BCE, and Urville-Nacqueville necropolis North Coast, 120–80 BCE, yielded only R1b. It does not look like I2a1b1a1a-M284 (ISOGG 2019) was expanding from France during LBA and IA.

Moreover, I had a look at the yDNA results of the upcoming DNA paper covering ancient DNA from France as they were published on Anthrogenica by an amateur, see

There are no I2a1b1a1a-M284 samples among these French results. The period of the samples is of course still unknown. The closest I2 lines are these: I2a1b1b2 and I2a1b1a2a1b2. They are probably Neolithic. IMO, it is more speculative to argue that the post-Neolithic I2a1b1a1a-M284 on the British Isles is from France or Germany if this line is hardly or not present in any continental post-Neolithic ancient samples.

Rob said...

@ Sam

'Development' goes beyond GDP. In the 'good old days', Yugoslavia in terms of literacy, age expectency, was better than many parts of USA are today. It stood out from other communist block countries after the Tito-Stalin split.

@ Kristiina
I prefer the old nomencleture. I2a2a1a has been found in the Wartberg culture & British Neolithic (more or less syncrhonous). A couple of Bronze Age Hungarians have it too (Vatya, Mokrin cultures). It is interesting that it seems to have "skipped" over France.
It presence in BA Hungary could be a West -> East diffusion (? British tin) during the Bronze Age

Simon_W said...

@Targamos the Based, son of Kavkasos son of CHG son of

"Off the top of my head Albanians are pretty similar to Montenegrins and Serbs genetically yet are way shorter."

Yes, in fact one of the best Swiss football players with Kosovar Albanian roots, Xherdan Shaqiri, has been called "power dwarf" or "power cube" by Swiss tabloids.


"It looks like they've just embargoed the thesis until 2023."

Apparently they read Eurogenes and Anthrogenica, otherwise I can't explain that erratic behaviour.

Wise dragon said...

@Samuel Andrews,

On anthrogenica, one user found out that the original Roman paper, Antonio et al. 2019 had Denmark_BA as 'potential incoming sources identified by qpAdm modeling', which gave rise to the Iron Age Romans. However, Denmark_BA was removed from the published Roman study. So the question is why? There was already a discussion about the Roman paper when it came out and the strange use of poorer fitting source population for their admixture model like Morocco_HG instead of Levant_N. The thing was that Morocco_HG strongly inflated the North African component. Plus the study used Iran_N instead of CHG. From what I recall, you sometimes pointed out that some studies appear to be biased in their conclusions. Overall what is your opinion?

Romulus the I2a L233+ Proto Balto-Slav, layer of Corded Ware Women said...

Neolithic British and Irish Y-DNA is far from extinct. I can say this for a fact, I'm part of the I2a P37.2 project on FTDNA and those subclades are plentiful, in particular I2a1b-L161.1. This subclade has been known about for a very long time.

Haplogroup I2a1b-L161.1
Commonly known in genetic genealogy circles as I2-M423-Isles, L161.1 is found at highest frequencies in western Ireland (5-10%) and the Scottish Highlands (1-5%), but is also found at low frequencies (> 1%) throughout Central and Western Europe, from Latvia, Lithuania and Belarus to the British Isles, and from Scandinavia to north-western Spain. It has also been found in Albania, northern Greece, Bulgaria and Romania.

British Isles I2a2 as well is Neolithic derived

As expected, Mesolithic samples show more basal lineages with respect to the Neolithic cohort, a significant number of whom placed within the subclade I2a2a1-M284, found almost exclusively in Britain today (International Society of Genetic Genealogy 2017; FamilyTreeDNA 2017). ***A near complete*** turnover is then witnessed in the Chalcolithic and Early Bronze Age, with virtually all samples from this time onwards belonging to the R1b-L151 haplogroup, associated with the Atlantic Modal Haplotype. This lineage forms a downstream branch of R1b-M269, the most prevalent haplogroup in western Europe today (Myres et al. 2011). It was possible to place a further majority of samples (81.8%) into the subclade R1b-L21, a haplogroup whose distribution shows a steep and somewhat restricted peak in modern Ireland, where it accounts for almost half of male lineages (Myres et al. 2011). Only one Y chromosome examined fell outside R1b-M269 after the Neolithic period, belonging to an Early Bronze Age individual from the southwest, Killuragh1, who placed within I2a2a1-M284.

Most likely *all* of British I2 is Neolithic derived.

Simon_W said...

@wise dragon
I don't know any significant traces of the Nordic BA in Italy. So maybe they deleted Denmark_BA for this reason.

Samuel Andrews said...

@Wise Dragon,

Are you suggesting Antino 2019 used unpublished Denmark BA samples?

I don't believe they did. Because There are a few low coverage genomes from Bronze age Denmark. You can find them in David's G25 PCA.

Denmark BA could be used to model Iron age Italians. But, Iron age Italians did not have ancestry from Denmark. They had ancestry from Bell Beaker in Germany.

Antino 2019 did a poor job identifying the source of STeppe gene flow in Iron age Italy. It is obviously mainly Bell Beaker. Which is key evidence Italic language is from Bell Beaker. Antino 2019 could have made a big deal about this. It is important to Indo European discussion.

But, I guess authors of Antinio 2019 didn't because they don't know a lot about the Indo European stuff.

Samuel Andrews said...

@Wise Dragon,

I don't believe reseachers are bias. What it is, is ancient DNA is a new and quickly changing field of study. They just can't keep track of everything going on.

Fregel 2019, who modelled Bronze age Sicily using IranN doesn't know there are Bronze age genomes from Near East which are much better references such as Kuras Araxes.

A said...

Or maybe they thought it was more important to emphasise the diverse multicultural Middle-Eastern nature of Roman civilisation rather than dwell on the fact that the Romans came from Germany.

Samuel Andrews said...

I seriously don't think the authors were political at all.

We need more ancient DNA from Rome/Italy. Historians already knew millions upon millions of forigeners (all were slaves originally) came to Rome. This is nothing new.

Btw, most of them have mixed Greek and Near Eastern ancestry. SOmething the paper didn't pick.

The DNA samples from Imperial Rome in the paper I don't believe represent Rome. I think the burial sites miss represent Rome. One burial site had carvings which literally said the people are descendants of slaves (therefore not native to Italy).

The Antiquity DNA samples are mostly Italian. Several samples are 100%, so the Latin/Italic population disappeared. But what exactly happened, how did they intermarry with immigrants, how did language and ethnicity in Roman era Italy. Lots of questions.

Samuel Andrews said...

If we had Juvenal's DNA, I'm sure he would turn out to be a true Roman/Latin unlike the dirty foreigners Antino 2019 got DNA from Lol.

"“The true born Roman whose infancy has drunk in the air of the Aventine and was nurtured on the Sabine berry must yield place to such as these.”"

Admium said...

Rome was already overflowing with "immigrants" by the first half of the 2nd century, as recorded by Livy. Granted most of these were probably from the Italic allies, but they were outsiders nonetheless. Add to that the Roman practice of cremation and the ever increasing number of exotic slaves and it's going to be really hard to tell who exactly the actual Romans were.

Wise dragon said...

@ Samuel Andrews,

It wasn‘t me that discovered the Denmark BA and Denmark LBA from the Roman paper but one user from anthrogenica. Here is the link:

Well, I don‘t think that scientists are(totally) without bias. It's an open secret that many in the academic world do lean very much to the left. Having said I hope that they don‘t let their bias interfere with their work. However, you can see the bias of the scientists, researchers by their use of language. In the unpublished study, they mentioned the BA and LKA folks from Denmark as a potential incoming source who gave rise to the Iron Age Italics and the map showed some ancestors from the Italics to come from Germany or Central Europe. In their published version they replaced them with Yamnaya from the Pontic Steppe and kept silent about the Bell Beaker or Germany as an ancestral group of the Italics. Not many outside the archaeogenetic community know where the Pontic Steppe is. When it comes to referring to the Anatolian farmers they can't stressed enough that they came from the Middle East. Asia minor would be a better term since people associate the Middle East with Arabs. Plus the Cheddar man hysteria taught me to take genetic studies with a grain of salt. For instance, the involved researchers that reconstructed the Cheddar man used the darkest skin tone possible and talked about the Cheddar man having the very dark skin of Africans, etc. Hence they started the black Cheddar man hysteria all over the internet and the media. Don't get me wrong there is nothing wrong with the Cheddar man being dark-skinned or Rome being full of foreign people and "diverse", etc. The point is it's wrong to politicize history or genetic findings to push the liberal and diversity agenda of TODAY. That bothers me since politics and ideology should be strictly kept out of science.

Besides, Eastern Meds were cool and the Byzantine Empire great.

Simon_W said...


In case Domen finds the time to send you the Alemannic samples from Niederstotzingen (and if the overlap enough with the G25 markers), I think it would be sensible to split the DEU_MA sample into DEU_Baiuvari_MA and DEU_Alemanni_MA, with ACD/_o subgroups respectively.

Simon_W said...

Imho it's easy to tell who the actual Romans were. ITA_Rome_Latini_IA RMPR1016 on the G25 sheet is from the Republican Rome IIRC. And he resembles other Latin samples from that time. Chances that the real Romans were something different are practically 0.

Simon_W said...

Nor do I think the real Romans/Italics have disappeared. Even before ancient Italic DNA was available, TreeMix models (I don't recall in which paper) suggested that central Italians are a mix of an Iberian-like pop with a Near Eastern-like admixture edge.

Simon_W said...

The real question is rather what the other Italics (other than the Latins and Hernici) were like. And on this the next ancient Italian DNA paper is hopefully going to shed some light.

Samuel Andrews said...

@Wise Dragon,

On, second look I see what you are saying. It does seem Antino 2019 authors intentionally ignored Germany as source of ancestry fro Iron age Italy.

I am also aware of the political bias that sometimes impacts ancient DNA papers. A big issue I have, is how they ignore PC Steppe is inside Europe. And, how they David Reich's lab calls it the Eurasian Steppe. Or how, often they say West Eurasian when they should say European.

Overall the bias is small in ancient DNA reseach, and facts always reign supreme in ancient DNA research.

The media reporting of ancient DNA reseach is where bias is biggest. The cheddar man debacle was all media. It was not done by researchers. Also, British media is silent on Bell Beaker arguable the biggest discovery of British ancient DNA. The reason is Bell Beaker reinforces the idea British are native to Britain.

Ryan said...

@Samuel Andrews - Yeah, your haplogroup must come from a Iron age clan who dominated that part of Ireland.

There's a Dal Cuinn facebook group and webpage that claims we're descendants of Conn Cétchathach. My own branch (the clan of Duibhiorma) I believe claims descent from Niall Noígíallach's son Eógan, down to the Lords of Inishowen (literally the "island of Eógan"). They were dispossessed around the early 17th century IIRC.

Look for "Mac Gleadra" and you've found me.

I'm not sure how much of it I believe. There do seem to be distinct clades for the different dynasties. They define their scope as R1b-DF104, which is a bit narrower than M222 in general.

Re: L21 - isn't Rathlin Island EBA L21? So it's been in northern Ireland for a long, long time.

Gabriel said...


Do you think it could have been a migration from North Italy? Some people say the Roman region was totally shifted towards East Med...

Romulus the I2a L233+ Proto Balto-Slav, layer of Corded Ware Women said...


Your comments seem like you don't understand the lower half of the Italian Peninsula, and Sicily, were colonized by Greeks before the expansion of Rome. Half of Italy would look like those Phocacian samples before Rome expanded. The imperial era reflects a blending of the entire peninsula + Sicily. Were there lots of Near Eastern slaves imported? probably, but I know that slaves would not be exclusively near eastern, i.e. lots of Gaulish slaves. Greeks in Magna Grecia were living in a post-Alexander world so would have already had more Near Eastern mixture with those people and slaves from that era as well.

Long story short it's a complicated scenario with many different people moving in different directions in different points in time. The assertion that Near Eastern slaves were exclusively responsible for the shift we see in Imperial Roman DNA is totally ignorant of the role of Magna Grecia, and many other events. Half of the early Roman men died fighting Hannibal, that had to have a profound impact on their makeup as well.

Ryan said...

@Sam The media reporting of ancient DNA reseach is where bias is biggest. The cheddar man debacle was all media. It was not done by researchers. Also, British media is silent on Bell Beaker arguable the biggest discovery of British ancient DNA. The reason is Bell Beaker reinforces the idea British are native to Britain.

Uh. Bell Beakers aren't native to Britain.

Samuel Andrews said...


When, I said Iron age Italy I meant Central-North Italy. I understand Southern Italy may have been different, I know it was colonized by Greeks.

"The assertion that Near Eastern slaves were exclusively responsible for the shift we see in Imperial Roman DNA is totally ignorant of the role of Magna Grecia"

The vast majority of the Imperial Rome samples have mixed Greek, Near Eastern ancestry. What do you think that means? I can't think of an explanation.

One would expect, Greek colonies to produce people with mixed Greek and Bronze age Sicily-like ancestry. Not people with Near Eastern, Greek ancestry.

So, where's the evidence they came from Southern Italy? Unless, Magna Gracia was inhabited by people with lots of Near Eastern ancestry.

What is for sure, is they had lots of recent Near Eastern ancestry. This is abudently clear looking at their mtDNA/Y DNA.

andrew said...

Worth noting that the Thesis is dated October 2017, just five months after the Bell Beaker Behemoth paper was released.

Simon_W said...

@Samuel Andrews

"Also, British media is silent on Bell Beaker arguable the biggest discovery of British ancient DNA."

It's the same with Swiss media and the recent Swiss LNBA paper, they're completely silent on it. They do like to report on new archaeological finds from time to time, that's presumably less delicate to them.

There lately was an exhibition in Berne named "Homo migrans", about how people have always migrated, also into and out of Switzerland. They topped it off with examples of Swiss people taking a DNA test. A rather lousy one at that. One ethnic Swiss got some Native American admixture, which is obviously just excess ANE, but they didn't clarify that. A journalist of the NZZ found the exhibition confusing. He wrote something along the line that on the one hand they suggest that we're all migrants and we're all mixed up, and e.g. people buried with Frankish or Byzantine grave goods needn't be Franks or Byzantines. Yet on the other hand they claim it's possible to identify tribes and peoples with the questionable concepts of population genetics.

Simon_W said...


"Do you think it could have been a migration from North Italy? Some people say the Roman region was totally shifted towards East Med..."

Well, the Treemix analysis I wrote about had North Italians on a different branch. And the Latins are Iberian-like, that would fit the model. It's true that the Imperial Age samples from Rome on average are closest to Greeks from Kos, Romaniote Jews and Cretans. But the question remains how representative this sample is for Imperial Age central Italy as a whole.

Ric Hern said...

How long time does a person and his/her descendants have to live in a region before they are classified as "native" to a region ?

Simon_W said...

@Romulus & Samuel Andrews

The Prenestini outlier dating to the Republican Age of Rome is closest to Italians from Basilicata, followed by Campania, Apulia and Abruzzo. So, in contrast to the Imperial Age samples from Rome he's not closest to modern East Mediterranean samples. I guess he may be an example for the geneflow from Southern Italy.

Drawing upon G25 models: MBA Sicily already had some Anatolia_BA admixture, about 15.8%. But no Levantine or North African admixture. LBA westernmost Sicily had less Anatolia_BA and just a trace of North African admixture. On the whole BA Sicily was still closest to Sardinians. Then came the Sicels from mainland Italy, soon followed by the Greeks in the east and the Punics in the West. Greek admixture must have increased the Anatolia_BA, and the Punics the Levantine and North African ancestry. But modern-day Sicilians are more Levantine in the east, with 6.8% Natufian vs. 5.6% Natufian in the west. North African Taforalt ancestry is still low, it reaches 2.2% in the west.

I wouldn't ascribe all Levantine admixture in Italy to slaves, given that even some Roman emperors were Syrians. We're dealing with a Graeco-Roman cultural area that encompassed the Hellenised Anatolia and Levant.

Samuel Andrews said...

My opinion is.....

Ultimately, it takes just one generation to be native. If you call a place home, you are native. But there's different kinds of native.

I'm native to America in the sense I grew up here, and my family has been here for generations. I'm not native to America in the sense I'm not Ameridian which is what matters when talking about population genetics.

Different ethnic groups have different degrees of nativness because some have longer more pure history. For example, Georgians are more native to Georgia than English are to England.

And Neighboring ethnic groups can have different degrees of nativness. Saami are more native to Northern Scandinavia than Swedes are even though both live there.

The idea of nationality is politically incorrect in Europe. Which, is why European mainstream only like to talk about DNA when it suggests their ethnic groups are not very native.

In regards to Britain, I mean c'mon yes British and Irish are native to Britain and Ireland. I don't need to explain why.

Samuel Andrews said...

I think there's unrealistically high expectations for nativness because of an evolution-based look at human diversity.

Evolution in a way has caused us to look at human populations like animal species/breeds.

Different animals are adapted to specific regions which is why they only live in specific regions and habitats.

Humans are not like this. Humans can live anywhere. Different human populations aren't biologically adapted to a region like animals are.

Chinese aren't adapted to an Asian habitat. Africans aren't adapted to African habitat (Dark skin blocks the sun, but that's not super important).

And, human populations are very willing to migrate whenever they see new better opportunities.

Hence, we shouldn't be surprised basically all human populations have mixed ancestry in the not so distant past.

If you look at Europe. Migrations didn't end with Neolithic farmers and Steppe pastorlists. Inter-Europe migrations occured after that. We know of several examples of migration within Europe in the last 5,000 yeas.

Most European ethnic groups mostly derive from Bronze age pops in their region. But all pops have ancestry from migration waves which occurred in last 4,000 years.

The Middle East, is a super inter-related region, because migrations have gone in almost every direction in the last 7,000 years. Even if they have little ancestry from outside Middle East going back 15,000 years, many migrations occurred within the Middle East.

But, despite those migrations which are rare, we certainly call people in native in very meaningful way. Not native like how limurs are native to Madagascar.

Ric Hern said...

@ Samuel Andrews

Humans did adapt and evolve in their regions. There are a lot of medical evidence for this....but yes Modern technology reduce the impact of natural selection.

Gaska said...

Regarding to Cassidy's thesis, we have to thank her for the work she has done because it is very complicated to rescue these ancient genomes. That said, as always, the conclusions of the thesis are a disaster, which is usual when a work is published without updating the available genetic data -for example Cassidy uses the data published by Rui Martiniano on Iberia 4 years ago when P312 only had been found in the Bronze Age (Post-2.000 BC). It seems that this lady is unaware that this lineage has been found in Spain 2,500 BC. No wonder they have embargoed the thesis, they have to update and review the conclusions because they are inaccurate and unprofessional.

However, there are some interesting data regarding the Irish Neolithic that confirm previous papers-Irish and Iberian genomes are indistinguishable (we only have to check the uniparental markers because they are identical)- There was therefore a Mediterranean migration but evidently it could not have its origin in Iberia because those genomes are also shared by the French Neolithic cultures, so the first Irish farmers were able to sail from Brittany. I think it doesn't matter much about the exact origin because the Breton, Iberian and southern French farmers were the same people- Archaeologically we have the evidence of Western megalithic culture

Regarding the Bbs, it is a pity that almost all genomes are from the bronze age, but uniparental markers tell us about a first Iberian migration (there are two female markers exclusive to Iberia) probably between 2,400-2,300 and related to mining on Rathlin Island and a second more important migration from Great Britain. These second early bronze age (Post 2000 BC) Bbs migrants originated in Germany, not in the Netherlands because there are many female markers that coincide with the last period of the bb culture and the beginnings of the Unetice culture in Germany. Then, for me, it is clear that our Irish friends are a bit Iberian and very German and that the British were in charge of spreading the German genes in Hibernia -This also confirms what many English archaeologists have said about the influence of the German BB culture in Great Britain (pottery styles etc)

Simon_W said...

@Samuel Andrews

Swiss mainstream media have little problem with calling the Neolithic pile dwellers or the Celts our ancestors. I think the scandal that comes with population genetics is the idea that DNA can tell something about ethnic affiliation, or in other words, that there's a biological/genetic aspect to ethnic groups. Especially in Germany this is an absolute no-go. That's a counter-reaction to the Nazi obsession with racial purity.

Simon_W said...

That's why so many leftist liberals have a problem with Zionism and hence with Israel. Because Zionism is the idea of creating a state for all people of Jewish descent, no matter what language they speak or if they believe in Judaism at all.

gamerz_J said...


"Greek admixture must have increased the Anatolia_BA"

Not entirely sure, how much Anatolia_BA did ancient Greece have? It's hard to calculate imo because the ancestral components are similar.

I doubt that populations from Magna Grecia are solely responsible for the Near Eastern shift observed in the Imperial samples. Simply because the Imperial samples have too much Levant_N and some North African that I don't think Greek-speaking populations had.

gamerz_J said...

@Samuel Andrews

"Fregel 2019, who modelled Bronze age Sicily using IranN doesn't know there are Bronze age genomes from Near East which are much better references such as Kuras Araxes"

I think they are more interested in modelling the deep sources of ancestry in these populations and since Kura-Araxes had Iran_N and CHG ancestry (closely related populations) then it seems to be good enough for them, even though it of course was not directly an Iranian Neolithic migration to Sicily.

Gaska said...

Although they are overwhelmingly L21 and we are overwhelmingly df27, Irish and Basques share certain genetic and cultural (Catholicism) peculiarities that make us a little different from other Europeans - we are both the most legitimate descendants of Western Bbs, our genomes are very similar to those first p312 of 4500 years ago (maybe that's why I have dozens of Irish matches) and we also share to some extent Neolithic genomes-Perhaps because we are peoples who have been very isolated genetically?-

The Basques and the BBculture never spoke an IE language and the Irish if we have to pay attention to their genetic make-up did not either. I am not an expert in Irish history and I don't know when the Celtic language came to the island or what its origin is, but it certainly had nothing to do with the BB culture. You will have to study your Iron Age to find the solution to the language issue - if the Irish want to hear how their ancestors spoke they just have to wait for the Chinese motherfucker virus to pass and spend a weekend in San Sebastian

J.S. said...

7,000years of demographic history in France

A team led by scientists from the Institut Jacques Monod (CNRS/Université de Paris)1 have shown that
French prehistory was punctuated by two waves of migration: the first during the Neolithic period, about 6,300 years ago, the second during the Bronze Age, about 4,200 years ago. This study, published in PNAS on May 25, which looked at the genomes of 243 ancient individuals over 7,000 years, demonstrates how admixture between native hunter-gatherers and the first Anatolian Neolithic migrants, who brought with them a lifestyle based on agriculture, persists to this day in the genomes of French people. Admixture of the Neolithic populations with those from the Pontic steppes2, who arrived 4,200 years ago in what is now France, also left a lasting imprint, with the Y chromosome of the majority of French men still bearing the signature of men from the steppes.

GregH said...

Does anyone have a link to the study this article discusses or is there no preprint available yet?

"The discovery of the burials of four medieval knights near the Polish village of Cieple has highlighted the region’s connections to Scandinavia during the reign of the first Polish kings"

"Isotope and DNA analysis demonstrated, though, that these individuals were not locals, but instead likely immigrated from an area around Denmark. The four warrior tombs were found at the center of a necropolis that contained at least 60 other individuals, of both local and Scandinavian ancestry"

"The men were likely members of an elite group of riders that ruled part of eastern Pomerania—present-day northern Poland—perhaps on behalf of the Polish kings, according to archaeologist Sławomir Wadyl of the Archaeological Museum in Gdansk"

A said...

According to this:

in the middle of the first century AD, an estimated 55% of the population of city of Rome were citizens, whilst 15% were noncitizen immigrants and 30% were slaves. Across Italy 70% were citizens whilst 5% were noncitizen immigrants and 25% were slaves.

In contrast, in Greece, Asia Minor, other Eastern Provinces and North Africa, only 1-3% were citizens, 70-80% were noncitizen residents, and 17-28% were slaves.

Does anyone know if this is accurate or not?

Vladimir said...

Vladimir said...

Supplementary Information for
Ancient genomes from present-day France unveil 7,000 years of its demographic history.

andrew said...

@RicHern "How long time does a person and his/her descendants have to live in a region before they are classified as "native" to a region?" It entirely depends upon what you want to learn by asking the question. Operationally, modern DNA studies usually look for four native born grandparents.

@Gaska "That said, as always, the conclusions of the thesis are a disaster, which is usual when a work is published without updating the available genetic data -for example Cassidy uses the data published by Rui Martiniano on Iberia 4 years ago when P312 only had been found in the Bronze Age (Post-2.000 BC). It seems that this lady is unaware that this lineage has been found in Spain 2,500 BC. No wonder they have embargoed the thesis, they have to update and review the conclusions because they are inaccurate and unprofessional."

That isn't how it work. You publish a PhD thesis once. In this case, in October 2017. Then, it's over, a done deal. When new information makes a past publication's conclusions doubtful, you don't update the old paper, you write a new one and publish that. There is nothing unprofessional about failing to foresee information that doesn't exist when you write it. The Embargo has nothing to do with the reason you suggest, and instead, is probably designed to protect the value of the work to scientific journals that don't allow preprints.

AWood said...

I suspect that the change from G2a/H2 to I2a occurred in Spain as a Spanish paper suggested a few years back. I think you will find these I2a lineages are common in Neolithic Spain as well. Therefore, by the time the Atlantic Neolithic hits France and then traverses to southern Britain -> Ireland, most of the lines are I2-M423 (Loschbour), I2-M284, I2a(xM26). I believe the Cassidy paper suggested one Neolithic Irish sample is very WHG like, near a swampy area that had evidence for late foraging, which suggests some Mesolithic communities may have been absorbed by the farmers. However, due to the fact the local foragers were highly inbred and in small numbers, this was most likely not the case for the most part, this particular example being an exception perhaps. All evidence points to a outbred, thriving Neolithic community in Ireland, as per the evidence presented through the duration.

What's also interesting is that this suggests all non-L21, non-I2 (subclades mentioned above) are probably middle-late Bronze Age, or later in UK/Ireland. This means I1, I2-Yamnaya/Continental European, G2-P303, J2, R1b-DF27..etc are late arrivals.

AWood said...


If it's any consolation, I believe the British/Irish Neolithic is entirely Iberian, you can see this in their Y chromosomes. However, this does not include R1b, so clearly Iberia has nothing to do with its ethnogenesis.

Anonymous said...

When his ancestors have participated in every mayour event that have formed the country.

10/ 10 american pionners
9/10 ancestors that fought in the independence war
8/ 10 american civil war
6/ 10 ancestors that fought in the ww1
5/ 10 ancestors that fought in the ww2

I cant consider the post ww2 inmigrant americans

Gaska said...

@Andrew, a doctoral thesis that is not published for three years only has the value of providing the results obtained, but its conclusions are obviously not of scientific value. Perhaps this is common practice in Ireland but in any case it is an absurd practice. If you then finally publish it and immediately withdraw it for another three years we can only think that something strange is happening. If in the year 2023 the conclusions of Cassidy's thesis are the same as those of 2017, you will be right, nothing will have been changed, but you will agree with me that it will be a waste of time to read it

Gaska said...


We have over 1,000 Bronze Age genomes in Europe and DF27 has only appeared in Iberia (17 samples) - this lineage is not in Ireland, England, Germany, France, Italy, Greece or Poland-I know this is going to be a trauma for many people but you can start thinking that Df27 on the islands probably arrived with the Iberian military units that participated in the Roman conquest-The anti-Iberian band will need psychological treatment

Let's see what happens with R1b-L51, we already have ATP3 (3,389 BC) with a small percentage of steppe DNA, 3 Swiss Neolithic farmers who have nothing to do with CWC, and Rivollat is going to publish a Neolithic R1b-L2 genome. We will have to wait to check the BAm file and the dating of this sample but if it turns out to be P312 that will definitely be a game changer-Many people have to abandon the paranoia of the Aryan conquerors who swept through Western Europe and apologize for their stupidity

Regarding L21 its origin is still a mystery, it has not appeared in continental Europe because in Germany, Bohemia, Switzerland, Poland northern Italy etc there is only U152. I don't know what will happen in France but in Spain we have a sample in Burgos (2.434 BC) with a positive SNP downstream L21-

Rob said...

@ AWood
How did you come to such a conclusion ? It goes against the evidence
The I2a1b and I2a2a in British Neolithic point to France & the Rhineland, resp
These same people migrated into Iberia during the middle neolithic and pre-BB copper age.

The native Iberian foragers are evidently C1a & I1

Jatt_Scythian said...

Slightly off topic but did we ever confirm if the blue dots here are modern and not ancient samples?

Ryan said...

@Srtmil - what?

Mr. J said...

There is definitely more at play than just genetic height and WHG ancestry when it comes to stature among modern day Europeans. Nutrition is another big factor I think. The very tall average height for the Dutch is related to nutrition surely since until fairly recently the average Dutch standing male height was not all that tall. The Western Balkans on the other hand was always overall a very tall region and at least in Montenegro there has been selection for taller stature. There was likely selection for taller stature in other areas there in the past as well. Traditionally pastoralism has been a major livelihood among West Balkan populations so this type of diet likely contributed to some degree as well.

On the other hand now there are other population within Europe such as the Sardinians and Basques that have large amounts of WHG ancestry and much shorter average stature. Shepherding is also important in those regions yet they both are still on average not all that tall.

Add to all of this it is known that the Yamnaya type peoples had Gene's for taller stature I think. In the end stature is a complex trait in which multiple factors interact I believe.

Samuel Andrews said...

That I1 was miss reported. It is I (xI2a2, I2a1)

Samuel Andrews said...

Davidski can you plz just ban Gaska already.

A said...

San Sebastian looks awesome.

Davidski said...


These samples are already in the Global25 datasheets.