search this blog

Saturday, June 27, 2020

Major updates to ADMIXTOOLS


An important message from Nick Patterson:

Dear Eurogenes bloggers,

Many of you use ADMIXTOOLS and you might like to know that there is a new release on github [LINK] with some important enhancements.

From the README

*** NEW ***

1)

Version 7.0 has numerous upgrades.

a) Two new executables --qpfstats qpfmv allow precomputation of f-statistic basis. This can greatly reduce computation costs.
b) qpAdm, qpWave, qpGraph support qpfstats output as input.
*** This is a much improved way of running with allsnps: YES. ***
c) A new experimental feature of qpGraph (halfscore: YES) allows comparison of 2 phylogenies + a (weak) goodness of fit score. Be careful if running with a large number of populations and consider reducing block size say blgsize: .005

2)

Note that several of the new ideas implemented in version 7.0 were developed collaboratively with Robert Maier, who has implemented them along with the great majority of other ADMIXTOOLS functionality in R: See https://github.com/uqrmaie1/admixtools
Executables run fast, and it has features not available in this C version, such as interactive exploration of graph phylogenies.
A manuscript describing the algorithmic ideas and providing documentation of the methods is in preparation.

qpfstats is the most important new executable. This estimates f-statistics and covariance on a basis.

a) This can be passed into other programs of the package without having to reaccess the genotype files, greatly speeding the computations.
b) In allsnps: YES mode a new computation is carried out (explained in qpfs.pdf) that is much more logical when there is a lot of missing data. Sometimes standard errors are greatly reduced.
qpfstats can be used with up to 30 populations. Much beyond that the output files become large.

As usual there may be bugs...

Nick Patterson 6/27/2020

Update 29/06/2020: As pointed out above, qpfstats is the most important new executable. Indeed, Nick Patterson now recommendeds that qpAdm analyses run with the allsnps: YES flag should be based on qpfstats output.

Several of my recent blog posts featured qpAdm models run with the allsnps: YES flag, but they were based on genotype data because obviously I didn't know anything about qpfstats at the time.

So I went back and ran some of these models again, just to make sure that they were still relevant. Below are three examples which you can compare to the original analyses here, here and here, respectively.

TUR_Arslantepe_LC_Maykop
RUS_Maykop_Novosvobodnaya 0.281±0.042
TUR_Arslantepe_LC 0.719±0.042
chisq 10.923
tail prob 0.449752
Full output

TUR_Barcin_C
RUS_Vonyuchka_En 0.137±0.031
TUR_Buyukkaya_EC 0.863±0.031
chisq 15.074
tail prob 0.0889099
Full output

UKR_N_admixed
RUS_Progress_En 0.083±0.020
UKR_N 0.917±0.020
chisq 6.825
tail prob 0.65538
Full output

As far as I can tell, they're very similar to the original runs, which is a relief, because it means that the conclusions in my blog posts still make sense.

52 comments:

Samuel Andrews said...

I'm going to start using ADMIX Tools soon. It is very cool amateurs can use the same tools experts use.

However It seems like a lot of ancient DNA authors aren't experienced in it and make beginner mistakes.

Cy Tolliver said...

"qpfstats is the most important new executable. This estimates f-statistics and covariance on a basis."

What exactly does this mean? And what exactly is the difference between F and Dstats? From my understanding F stats are for computing admixture proportions of populations (but isn't that also what qpAdm and qpGraph are for too?), whereas as D stats are used to test if there's been admixture or drift-sharing between sets of populations. But I'm pretty sure I've seen one-off F4 stats being run as if they were just normal D tests, so I'm not sure if there's an actual computational differences between the two methods?

Also, why does Admixtools only work on Linux systems? What's wrong with Windows?

vAsiSTha said...

Nice

Jatt_Scythian said...

Speaking of admixture I was going through some old posts on this blog.

http://eurogenes.blogspot.com/2018/03/siberian-ancestry-and-y-haplogroup-n1c.html?commentPage=2

So do Balts have Siberian ancestry or not? How much do Finns, Saami, NW Russians and German have of this admixture?

Also I see models for Iran_N as 20% East Eurasian, 10% AASI, 20% ANE and only 50% Basal + West Eurasian? Is that accurate? I figured Iran_N was pretty close to CHG and given there is CHG admixture in Europe without any excess ENA/AASI how does this make sense?

CHG Chad said...

@ Jatt_Scythian

I used to believe the same,that CHG and Iran N are the same components but now i have my doubts.I think Iran N is more eastern(i am not sure how accurate this 10% AASI) can be, but it contains more ANE and for some reason more Basal compared to CHG witch seems to be by far more western eurasian(maybe more Dzudzuana?).We have a J lineage in Karelia HG and the CHG component in Yamnaya and later steppish cultures shows definitely a west eurasian affinity.There are rumours and more specific about Onge related admixture in Iran N as well(thought we should wait and see if its accurate).Anyway i know this out of the forum rules and David might delete the comment(i dont have any rascist/superior motivation here) but Balochi populations 'phenomically' are not identical to populations with high CHG admixture(Georgians,LaZ,,Trabzon,Abkhazians,North Caucasians etc).So,the Iran N and CHG are not identical as we used to believe in the past.Yes,they share similar components but their origins are too complex to be identical to each other.

Jatt_Scythian said...

@Johnny Ola

I'd love to talk more about this. Can we email or discord? Thanks.

Onur Dincer said...

@Johnny Ola

Anyway i know this out of the forum rules and David might delete the comment(i dont have any rascist/superior motivation here) but Balochi populations 'phenomically' are not identical to populations with high CHG admixture(Georgians,LaZ,,Trabzon,Abkhazians,North Caucasians etc).

Not all of the populations you put in parentheses are genetically or phenotypically close. Also, the Baloch have about %15 Paniya-type South Indian ancestry besides Iran N according to my analysis, which makes them darker. Here are some population average results:

https://docs.google.com/spreadsheets/d/1-dExi7dxWVfM5p_fppGYRyNBXx0SBeBIDE0r78MpqKw/edit#gid=0

The Yagnobis have high Iran N, no CHG and almost no Paniya, and many of them have light pigmentation:

https://www.google.com/search?q=%22yaghnobi%22&tbm=isch&ved=2ahUKEwj086XG-6XqAhVK3hoKHX3XBakQ2-cCegQIABAA&oq=%22yaghnobi%22&gs_lcp=CgNpbWcQAzIECCMQJzICCAAyAggAUKfjA1iD-QNgz_8DaABwAHgAgAGVAYgBpgKSAQMwLjKYAQCgAQGqAQtnd3Mtd2l6LWltZw&sclient=img&ei=QFD5XvTTE8q8a_2ul8gK&bih=625&biw=1366

CrM said...

I think that Hotu/Iran_N is genetically closer to the more archaic ancestor of CHG, and CHG being 50%+ archaic Iran_N plus additional ANE/EHG and Anatolia/Dzudzuana. While Hotu/Iran_N is predominantly of that archaic Iran_N descent but with some additional ANE, Natufian and Onge-related.
The ANE-related ancestry I think was picked up by CHG and Iranians somewhere closer to the South Caspian?

CHG being just Dzudzuana + ANE doesn't make sense archeologically because I'm not aware of any Caucasian material cultures that could be related to ANE, rather all foreign invaders to the Caucasus came from the south. First it were the Gravettians, with their lithic culture being present in Ortvale Klde in Georgia, this is also discussed in the Gravettian paper, then came the Dzudzuana and then CHG. The arrival of CHG correspond with the movement of South>North cultures, those being Baradostian>Zarzian>Trialetian>Imeretian, last one being the culture of Satsurblia and KK individuals.

Davidski said...

@All

Please check out the update. I posted three examples of qpAdm models based on qpfstats output.

Thoughts?

vAsiSTha said...

I'll try understanding this and running some models in next week.

CHG Chad said...

@Onur Dincer


Balochi people indeed have a decent Paniya like admixture but the main component is Iran N.And ofc they do not look exactly like Paniya folks.As for Yagnobis,how much of steppe ancestry they carry?Something is telling me that these folks have eastern iranic nomadic tribal roots.

CHG Chad said...

@ Jatt_Scythian

Buddy,i wish i could talk with you,but i am working almost every day..(summer season in Greece) :( i have limited time to almost none.Even some of my posts both in Anthrogenica and also here are posted in breaks or at late night when i am off xD.I can only chat.

You can add me here if you want.

Ioannisg95@gmail.com

This is my Skype as well:Ioannis Gavras

You can add me for more chat there!!!

CHG Chad said...

@David

David plss can you make a model as well for Ikitzepe?The site has definitely some association with Hattians-Hittites.

Thank you!!!

Onur Dincer said...

@Johnny

Balochi people indeed have a decent Paniya like admixture but the main component is Iran N.And ofc they do not look exactly like Paniya folks.As for Yagnobis,how much of steppe ancestry they carry?Something is telling me that these folks have eastern iranic nomadic tribal roots.

The Yagnobis have about 40% Yamnaya-type steppe ancestry according to my model, and their Iran N is close to 30% according to the same model, the rest of their ancestry is mostly Anatolia N. Unfortunately there is no modern population who have close to pure Iran N ancestry, so our knowledge about the pigmentation of Iran N is very limited. The pigmentation genetic loci-based predictions on this page can give you a rough idea (search Iran EN):

https://genetiker.wordpress.com/pigmentation/

CHG Chad said...

@Onur Dincer

Their 40% Yamnaya like steppe ancestry gives them those looks.Balochis are the best example of Iran N look IMO despite their Paniya like component.A 15-20% is not enough to make a whole population look different.

https://genetiker.wordpress.com/pigmentation/

I have seen it before.But as a Turk yourself do you really believe modern Caucasians especially Georgians,Abkhazians,Laz( i am not putting north caucasian folks here for being heavily steppe admixed) have the same pigmentation with people from Balochistan?With a exception their very long and hooked noses(i am wondering if this is an ANE trait) i dont see other similarities.

Matt said...

I guess the precomputation of fstats saves system memory, esp under the allsnps condition where memory updating would be more frequent (because computing changing subset of SNPs)?

It doesn't seem to change too much in output. So far. The models with mid end robust p value are still robust, and the lower end p Barcin C model is still in the same general neighbourhood with the same worst vendettas stats (generally suggesting real Barcin C less related to EHG / ANS than model with Vonyuchka+Buyukkaya).

Onur Dincer said...

@Johnny

Of course it could be argued for the Baloch (and genetically similar peoples like the Brahui and Makrani) that their steppe and and Anatolia N ancestries neutralize the effects of their Paniya-like ancestry on pigmentation. I agree with you about the probable darker pigmentation of the Iran N than the CHG, the pigmentation loci results point to that as well, we just cannot quantify it in an accurate way with the data we have.

Davidski said...

@Matt

Yeah, this is better, but like I already pointed out, it's anachronistic.

https://drive.google.com/file/d/1bfKMqIKjksBdf3QMoLK_aAVj1Q5URptP/view?usp=sharing

Any suggestions?

Davidski said...

@Johnny Ola

David plss can you make a model as well for Ikitzepe? The site has definitely some association with Hattians-Hittites.

They look more Hattian than Nesa/Hittite to me, because the Ikitzepe LC people are Anatolians with an eastern, Caucasus-related genetic shift.

It's often said that the Hattians were related to Caucasian speakers and came from that direction.

Only one of the Ikitzepe samples might have a bit of steppe ancestry: IKI017.

CHG Chad said...

@ David

Ty. Yes i will agree with you that Ikitzepe seems to be Hattian related.Hittites are much later. As For 017 i think a Proto-Anatolian IE relation maybe of Palaic origins. Do we Have any idea about the exactly period that Anatolians started to migrated to North Anatolia? They might arrived in LC period hence the Steppe shift in 017 individual.

Onur Dincer said...

Since the CHG/Iran-related ancestry that spread in Anatolia and parts of Southern Europe beginning from the Late Neolithic was something between CHG and Iran N, unlike the one in the the steppe, which was heavily CHG-like, it should have a more southern origin, probably in the area stretching from Greater Armenia to the Southern Caucasus.

https://docs.google.com/spreadsheets/d/1a5byEejcptHST6X2L6Exb60uGXnyI6tZYI7HWG-EkkE/edit#gid=0

Matt said...

Not really. Considering only the stats f4(Real,Fit;Yamnaya_BGR,pright), I would have thought that you'd just need more Barcin_N related ancestry and less EHG related ancestry, so 3-way model of Buyukkaya+someBalkanN+Piedmont_En might work.

But the stat of f4(Real,Fit;Barcin_N,EHG) looks OK as is, so I am not sure about that.

Same thing is true with the model with Vonyuchka where the the stats f4(Real,Fit;Vonyuchka,pright) look like it wants far more affinity to Western farmers and less affinity to mainly EHG and also CHG.. but the actual stats of f4(Real,Fit;Barcin_N,EHG) is limited.

Quick graphics to show what I mean (red is stat where Real:Fit is Z > than 3 difference, green where Real:Fit is Z > than 2 difference): https://imgur.com/a/wVySik5

Maybe add in the Progress_En samples to pRight and try both models again and see what happens?

Davidski said...

@Matt

https://drive.google.com/file/d/14irR_SAdCJcFXBB2dYNvaWazj6cnVaOs/view?usp=sharing

https://drive.google.com/file/d/1_t8Ul3u1xxxMSDfqhmEd3llDAEW1f8eU/view?usp=sharing

Davidski said...

@Johnny Ola

There's a good summary of the various theories about the origins of the Nesa/Hittites/Anatolians here.

Encyclopedia of Indo-European Culture: Anatolian Origins

Interestingly, the theory that the Proto-Anatolians came from the Maykop culture has already been corroborated by ancient DNA, because the data from Skourtanioti et al. clearly show that there was admixture from Maykop, and even the North Caucasus steppe, in the Copper Age Arslantepe population.

Maykop ancestry in Copper Age Arslantepe

Of course, Skourtanioti et al. denied this, probably because most of the authors are from the Max-Planck-Institut für Menschheitsgeschichte (aka MPI-SHH), which is pushing a new theory that the Indo-European homeland was somewhere in the Near East, like Mesopotamia (haha).

In any case, my view is that the Proto-Anatolians came from the steppe and that they first moved into western Anatolia via the Balkans during the Copper Age, and I'm confident that this will be supported by ancient DNA.

Copper Axe said...

@Davidski

Aren't these statements a bit contradictory or am I a dumdum?

"Interestingly, the theory that the Proto-Anatolians came from the Maykop culture has already been corroborated by ancient DNA"

"In any case, my view is that the Proto-Anatolians came from the steppe and that they first moved into western Anatolia via the Balkans during the Copper Age, and I'm confident that this will be supported by ancient DNA."

Or are you saying that the Caucasian migration has evidence for it, but you don't think they were Early Anatolians and that they came from the west through the Balkans.

Personally, I keep flipflopping between the two. Maykop culture is really interesting, I think it would be cool if they were the Proto-Anatolians. However Luwians are a bit hard to link to this migration given their western location. Wasn't that R1b-M269 Phillistine likely to be Luwian or am I misremembering here?

"Of course, Skourtanioti et al. denied this, probably because most of the authors are from the Max-Planck-Institut für Menschheitsgeschichte (aka MPI-SHH), which is pushing a new theory that the Indo-European homeland was somewhere in the Near East, like Mesopotamia (haha)."

So Iran didn't work out and now they are going for Mesopotamia LOL. Is this how Germany tries to rectify their mistakes of the second world war?

CHG Chad said...

@ David

Do you have any clue what is the purposes about IE homeland outside from the steppe?Why these guys trying to prove the opposite?How they are going to convince us that IE's took part in Anatolia or in Mesopotamia-fertile crescent?I have no idea what are their goals and what exactly they trying to prove..but i am going to be a little bit political here and conspiracy theory 'moron'.I think there are groups behind it..like lefties,MGO,Soro's puppets and stuff like that who trying to convince to people with
European origins that IE's took part outside from EU(aka immigrants and not natives).This tactic seems to help them probably with the current situation on immigration issue that we all seeing especially nowdays in EU.There is no other explanation why these guys trying so bad to convince for the opossite.Also for people who are going to call me rascist or xenophobic and stuff like that i have not European origins(i am actually Anatolian Greek) and genetically i am also Middle easterner but the whole subject with IE origins outisde from the steppe is getting annoying A.F not to mention that it hurts people with anatolian,levantine or mesopotamian roots.I am pretty sure people with such origins feel comfortable and they have nothing to be jealous.All the human beings should be proud for their origins end of story.Sorry for my drama but this thing with IE's homeland it makes me mad and boring.I don't want to discus about IE homeland in 2020 for gods sake...

mzp1 said...

Looks like a cool update making admixtools much easier to use. There is even an R package included which looks interesting, I had been using admixr up to this point.

Hint: If you want to run Admixtools but dont have Unix it maybe a good option to 'rent' a Virtual Cloud machine (Azure, AWS etc) as they are quite cheap to operate.

On a different note, I was wondering what happens if you take a list of alleles from a target population, and the equivelent from source populations, and then compare allele by allele to see which source populations match the target closest. Would this be a decent way to analyse populations? I tried this out with a simple program but the results are a bit 'funny'.

Also, I just saw that there is a sample from Uzbekistan, Zamanbaba_N, 4500BC, from the Narasimhan paper. It's low quality but I was wondering if it has been analysed much, as I dont remember hearing about it before.

Davidski said...

@Copper Axe

Aren't these statements a bit contradictory or am I a dumdum?

They're not contradictory, because the first statement just says that if the Maykop culture was Proto-Anatolian then this has now been corroborated by ancient DNA, but the second statement explains that I don't think the Maykop culture was Proto-Anatolian.

gamerz_J said...

@Davidski

I don't see a lot of difference in the qpAdm you posted there but I am not qualified to talk about ADMIXTOOLS since I've never used them so far.

However, I'd be interested in your opinion on whether you think they change any inferences about the Yamnaya for example.

gamerz_J said...

@Johny Ola

I don't think Greeks are Middle Easterners, they are closer to Balkan populations rather than most Middle Easterner groups.

Davidski said...

@gamerz_J

However, I'd be interested in your opinion on whether you think they change any inferences about the Yamnaya for example.

Nope. At least not my inferences.

CHG Chad said...

@ gamerz_J


Greeks from mainland are Europeans and their high steppe input witch is probably coming from Slavs,Vlachs and Balkan folks making them 100% European(some post-neolithic west asian input can be found even among them.. but nothing crazy).In my situation, as an Anatolian Greek with ancestry from modern Gumushane,Trabzon and Ordu i cannot considered as a European individual not only from genetics but also from geographical position.I consider my self as Anatolian-Caucasian.My yDNA is also very anatolian(found it in Arslantepe) but as David have mention and its true Arslantepe has Maykop-Kura Araxes influences so it is very likely my lineage to be associated with a transcaucasian lad.In general Anatolian Greeks not only those From Pontus region but also from Cappadocia are mostly of native origins with some decent Greek/Mycenean like ancestry.To understand the Greek ethnogenesis you have to read Byzantium history.It is the source for the Greek ethnos with christianity playing the major role.Modern Greeks are a mess genetically.Many ethnic groups with only one thing in common the Religion.

Copper Axe said...

@Davidski

Thanks, that makes more sense.

Where is the Max Planck Indo-Mesopotamia stuff coming from anyways? Was it hinted at in Skourtanioti's paper or something? Or is this new stuff they will be talking about soon?

Davidski said...

@Copper Axe

See here...

https://eurogenes.blogspot.com/2016/10/dead-cat-bounce.html

https://eurogenes.blogspot.com/2018/12/some-german-guy-once-said.html

gamerz_J said...

@Davidski

Thanks, certainly sounds like a relief!

@Johny Ola

also thanks for the reply and I see your point.

Samuel Andrews said...

@Johnny Ola,

Interesting you are Anatolian Greek. Because I used to assume all of Anatolia became Turkic Muslim. It is interesting there are still Greek-speaking Christians.

Trabazon Greeks, I see lack pretty much any Greek or Turkish ancestry, you guys are almost perfect intermediate between Eneolithic Central Anatolia and South Caucasus. You definitely have ancient roots in that region.

CHG Chad said...

@ Samuel Andrews

Hallo Samuel(nice name btw,my grandpa got it and everyone passes him for a Jew xD).

Yes Anatolian Greeks are Christians and even more Christians and believers compared to mainlanders because our roots goes back to Byzantium Empire and culturally we are the closest people to it (not that mainland Greeks are foreigners,i dont want to be misunderstood) but Anatolian Greeks represent Byzantium Empire and Orthodox Christianity better. There are muslim Pontians those from Trabzon for example (i dont want to go political again,but they have become muslims by violence and very tough ways from Ottomans.Genetically they are like Trabzon Greeks.Actually some of them are crypto-christians like many Armenians and Assyrians in eastern anatolia).But lets put these things out.Yes genetically people from Pontus are mostly a mix of BA Anatolia(Hattian-Hittite) and Kura-Araxes-Maykop with some Greek/Mycenean admixture.The samples from Trabzon are ofc more caucasus shifted and this has to do because the folks there are mostly hellenized Colchians.Before the Greek colonization eastern Pontus was inhabit by tribes similar to modern Georgians,Laz,Kartvelians hence their CHG shift.Ofc people from Trabzon are more Anatolian N related compared to Georgians,Abkhazias and even Laz.We have more BA anatolian ancestry and Greek/Mycenean witch shift us to east med spectrum.


https://en.wikipedia.org/wiki/Colchis
https://en.wikipedia.org/wiki/Colchian_culture
https://en.wikipedia.org/wiki/Kingdom_of_Pontus
https://en.wikipedia.org/wiki/Empire_of_Trebizond


And this is my surname.If you are not bored you can read..

https://en.wikipedia.org/wiki/Gabras

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

Modern Greeks don't get their Steppe ancestry levels exclusively from Slavs, maybe in part. You can see from the Ashkelon paper that these Philistine migrants with WHG admixture that existed in neither neolithic Greece, Mycenean Greece, or Minoans, show up in 1,200 B.C. along with an R1b-M269.

1200 B.C. is when the Bronze age collapse happened, potentially related to an explosive migration of people from the Urnfield culture with whom it overlaps.

https://advances.sciencemag.org/content/advances/5/7/eaax0061/F2.large.jpg?width=800&height=600&carousel=1

https://advances.sciencemag.org/content/advances/5/7/eaax0061/F3.large.jpg?width=800&height=600&carousel=1

CHG Chad said...

@ Romulus

Modern Greeks have steppe from many sources but slavs and vlachs are the main,and its also obvious from many Ydna lineages.Ofc they got some Post-Mycenean steppe admixture that No1 can doubt.As for the Philistine migrant with WHG,it can be easily a pirate from Italy,Sardinia,Illyria,balkans etc.It is well known that Sea people were a mess genetically..i am not rejecting the fact that we are going to found a 'Viking' in the upcoming 'Sea People' samples rofl!!!!

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

Regardless, these people didn't have a big impact as we can see by the Phocacean samples who were from 500 B.C., also in Ashkelon that dna disappeared quickly. AThe Imperial Roman DNA indicates an east to west influx of DNA into Rome during that period. Which I at least think is mostly from Magna Gracia.

Davidski said...

@mzp1

Also, I just saw that there is a sample from Uzbekistan, Zamanbaba_N, 4500BC, from the Narasimhan paper. It's low quality but I was wondering if it has been analysed much, as I dont remember hearing about it before.

Just had a quick look at this sample with a couple of PCA.

It looks to be contaminated with modern European DNA.

Nezih Seven said...

@Johnny Ola South Caucasians (mostly Kartvelian-speakers) and Pontics (Greeks & Hamshenis) are considerably different:

https://nezihseven.wordpress.com/2020/05/04/dna/

Gabriel said...

Don’t the Mycenaeans show (and are going to show even more) varability in steppe ancestry, like in Crete_Armenoi? I think that’s where a lot of Greek steppe ancestry comes from.

On the other hand, there’s the Empuries Greeks...

Rob said...

@ CRM

“ CHG being just Dzudzuana + ANE doesn't make sense archeologically because I'm not aware of any Caucasian material cultures that could be related to ANE, rather all foreign invaders to the Caucasus came from the south. First it were the Gravettians, with their lithic culture being present in Ortvale Klde in Georgia, this is also discussed in the Gravettian paper, then came the Dzudzuana and then CHG. The arrival of CHG correspond with the movement of South>North cultures, those being Baradostian>Zarzian>Trialetian>Imeretian, last one being the culture of Satsurblia and KK individuals. ”

Of course there is. - the epigravettian
Shouldered points from Eastern Europe

gamerz_J said...

@Romulus

" AThe Imperial Roman DNA indicates an east to west influx of DNA into Rome during that period. Which I at least think is mostly from Magna Gracia."

I don't want to derail the thread, but I find it very unlikely samples with high Levant_N ancestry to be coming from Magna Grecia. Not all Greek samples have this component, and is in fact quite low where present.

Joey said...

I would just like to give credits to Nick for being this connected with the amateur community.

Pavel Flegontov said...

Hello! I suggest that you also pay attention to another re-implementation of Admixtools, Admixtools2 in R, written by Robert Maier from Reich's group: https://uqrmaie1.github.io/admixtools/reference/index.html

The tool implements a wider set of functions as compared to Nick's version, and is also much faster than older Nick's versions.

On another note, in many recent papers qpAdm and qpGraph tools are used improperly. I'd be happy to talk to you about these methodological problems as soon as my preprint discussing these issues is published in 2 weeks.

Pavel Flegontov, Reich Lab

Davidski said...

@Pavel

Thanks, looking forward to it.

Pavel Flegontov said...

Hello! My preprint I've talked about is out today:
https://www.biorxiv.org/content/10.1101/2020.10.12.336628v1

The paper is devoted to the settlement of America, but it has a huge SI section on method testing and development, and includes improved qpAdm and qpGraph protocols and new methods. If you are interested in using these novel protocols, I'd be happy to talk via zoom or skype or email. There is a lot to discuss about biases etc.

Pavel

Davidski said...

Awesome.

I'll start a new discussion thread about your preprint tomorrow, and we'll go from there.

vahaduo said...

@ Pavel Flegontov

Both .html files from the supplement section do not render correctly.

Pavel Flegontov said...

@vahaduo
Strange, please try downloading them instead of opening them as a link in a browser.

@Davidski
Great, I'm looking forward to that.