Monday, 9 March 2020

A look at some new neolithic Irish genomes

Main passage tomb Listoghil, Carrowmore.
Samples were taken from 2km away in Primrose grange dolmens.
The individual buried here at Carrowmore 4 was deemed to be
the father of one of the Primrose males. 


With the recent publication of genotypes on the Reich lab’s Harvard website featured in last year’s paper on several neolithic samples from megalithic tombs across Northern Europe including several from Primrose and Carrowmore in County Sligo (you can read it here), I decided to take a look at some of them using the Global 25 tool courtesy of the eurogenes blog.
By far the best for modelling ancient populations.

Due to the lack of autosomal DNA samples from ancient sites in Ireland in the past few years, and with the Genomic Compendium of an Island embargoed for another two months. I decided to pass the time and boredom by playing around with a few of the Neolithic genomes featured in last year’s paper whose genotypes were only released quite recently. They’re not as high quality as the one shotgun sequenced ballynahatty sample from the 2015 Cassidy et al paper.
But beggars cannot be choosers I suppose.

In line with recent genetic findings the neolithization of Ireland was a demic event carried out predominantly by cardial agriculturalists spreading from the Aegean via a Mediterranean route eventually reaching the North Atlantic as opposed to the Danubian pioneer populations moving from a more direct Central European route. I wrote an article about which route Ireland was populated from last year concluding with the evidence it was the former and not the latter. The Danubian farmers were generally more devoid of WHG related ancestry as opposed to the cardial populations who assimilated the various Epigravettian and Magdalenian offshoot groups that lived across Southern Europe. The presence of mtDNA haplogroups in Irish neolithic samples that were common in cardial Iberian mitogenomes also provides further evidence (K1a,HV0,H1,T2)
Even more significantly all of the male samples belong to the same Y-DNA haplotype (I2a2a1a1) snowing evidence of patrilineality. The researchers came to the conclusion that these individuals all belonged to the same kindred that lived over the course of twelve generations, allowing thirty years per generation.


D-stats from Cassidy et al 2015 showing Irish neolithic samples sharing
 more drift with cardial related farmers


But how do we know Ireland was populated from the so called Atlantic facade for certain? How can we be sure this drift isn’t just some artefact from similar ratios of WHG admixture to cardial groups rather than actually coming from them? How do we know Ballynahatty’s population didn’t come from a Danubian route and assimilated some hunter gatherer bands along their way?

We know by taking a closer look at the HG component of the Neolithic Irish genomes, which seems to possess ancestry related to Magdalenian complex samples from Iberia rather than just plain old WHG from Central Europe. Magdalenian related ancestry provides a tracer dye for movement of Neolithic groups. They were the descendants of the first modern humans to reach Europe 40kya and sought refuge in Iberia during the LGM before eventually being assimilated by incoming WHG foragers towards the end of the Pleistocene. Both eventually being assimilated by anatolian farmers. 
Looking each of the six samples available to me this type of ancestry is self evident, for my model I will use early neolithic samples from Atapuerca (Iberia_N) and Stuttgart (DEU_LBK) respectively.
For their HG ancestry I will use the 19kya Magdalenian from El Miron.


Starting off with the most high resolution sample from dolmen 002,
the presence of Danubian farmer ancestry could be an artefact or
perhaps real authentic introgression from them did occur with
cardial groups in the Paris basin.


At least three of the Primrose grange samples can be feasibly modelled
sans any Danubian farmer ancestry and are completely Iberian derived
Iberia_N = 90% ANF + 10% WHG


Primrose 016 appears to be an outlier as he harbours roughly one third
of his genome from Danubian EEF, it is difficult to distinguish
such autosomally similar populations so I decided to make another
model just for him.


Despite the improved fit with samples from the Michelsburg culture in
the Paris basin he still requires the same proportion of Danubian
making him the purest Neolithic sample from this site.

Average of all six samples.


PCA revealing variation between Irish neolithic samples and contemporary populations from
across Europe at the time.
Primrose 12 appears to be equidistant between Sardinian and Basque clusters respectively.