PDA

View Full Version : The link between South Africans and Oceanian people



BalkanKiwi
10-19-2015, 10:03 PM
http://i.imgur.com/PR3pGjc.jpg?1


As some of you may have seen, I've been looking into the link between South Africa and Oceania. I thought it would be appropriate to start a thread about the topic.

To start off with, on paper I'm 0.8% Maori (an extremely small figure I know) however I score between 1-1.6% Oceanian/Papuan on most calculators. I assume, even though still a small amount, I've inherited more than expected. The value, big or small, isn't of significance to me at this stage, more so its connection to other populations.

What got me curious about is the trace amounts of Bushmen I get on the MDLP K27. It may be slightly noisy, but I like the components it contains. This led me to the small amounts of South African I get on the PuntDNAL K15, as well as the Eurogenes K9b and to a lesser extent, the Africa9 (noisy for Europeans). These are my results for the select calculators.

Eurogenes K9b

South_African 0.56%

Africa9

23andMe - San 0.53%
FTDNA - San 0.43%

puntDNAL K15

23andMe - S_African 0.35%
FTDNA - S_African 0.37%
Sister - S_African 0.59

MDLP K27

23andMe - 0.15% Bushmen
FTDNA - 0.19% Bushmen

I rarely score any other type of African i.e Sub-Saharan or Paleo African, and if I do, its 0.01-0.04% or slightly higher if the cluster included South African samples for whatever reason.

To me, this suggests the South African cluster is very specific.

With speaking to puntDNALKing, the creator of the calculator with the same name, I have learned that there is a clear indication that the South African component has contributed to the Oceanian component. With his permission, here is an image of the TreeMix for the PuntDNAL K15 calculator, which has a genotype rating of 99.8%. You can see a clear link between South African and Oceanian.

http://i.imgur.com/YSRsH8Q.jpg

Dienekes also reported something similar back in 2012 with the Globe 13 calculator. His TreeMix has a link between Paleo African and Australasian, as evident here.

http://1.bp.blogspot.com/-IS_fRlD7V5k/UJBi-JEMqjI/AAAAAAAAAzs/FWBI9YH6zLU/s1600/treemix.png

In talking with puntDNALKing, there are clearly new hypothesis that indicate there is more than one out of Africa theory. The new research showing back migration into Africa demonstrates more went on than what we currently know.

Someone who is more Polynesian/Papuan than myself would be interesting to look at with regards to what African components they score.

I would like to think the work done by Dienekes three years ago, and now with puntDNALKing, shows that there seems to be a credible link between South Africa and Oceania.

puntDNALKing
10-20-2015, 02:58 AM
Thank you for creating this informative thread Jtoml4. Although I read about this few months back, I never delved into this research. I found an article, here (http://www.nature.com/news/genomes-link-aboriginal-australians-to-indians-1.12219), that can shed some light on this; the article states what we already knew, that Oceanian popuations, including the mamanwa, have ancient DNA that split from the OOA movement 24,000 years before Europeans and Asians split from each other. However, the surprising aspect of this article, at least to me, was the finding that Australian Aborigines alone, not new guinea or mamanwa, have more recent admixture from South Asia, around 4,500 years ago. Now lets get to my main point; we know the original OOA people were based in East Africa, modern day Ethiopia. According to this article, here (http://www.ncbi.nlm.nih.gov/pmc/articles/PMC384897/), Khoisan territory once extended all the way to what is now modern day south Ethiopia and Sudan. Logic follows that Khoisan, who have the oldest DNA in Africa, diverged from the original Africans that give birth to the OOA population, thus that is the reason these two share ancient connection. This is purely speculation on my end, let me know what you think.

BalkanKiwi
10-20-2015, 03:38 AM
Thank you for creating this informative thread Jtoml4. Although I read about this few months back, I never delved into this research. I found an article, here (http://www.nature.com/news/genomes-link-aboriginal-australians-to-indians-1.12219), that can shed some light on this; the article states what we already knew, that Oceanian popuations, including the mamanwa, have ancient DNA that split from the OOA movement 24,000 years before Europeans and Asians split from each other. However, the surprising aspect of this article, at least to me, was the finding that Australian Aborigines alone, not new guinea or mamanwa, have more recent admixture from South Asia, around 4,500 years ago. Now lets get to my main point; we know the original OOA people were based in East Africa, modern day Ethiopia. According to this article, here (http://www.ncbi.nlm.nih.gov/pmc/articles/PMC384897/), Khoisan territory once extended all the way to what is now modern day south Ethiopia and Sudan. Logic follows that Khoisan, who have the oldest DNA in Africa, diverged from the original Africans that give birth to the OOA population, thus that is the reason these two share ancient connection. This is purely speculation on my end, let me know what you think.

You're most welcome. With regards to the first article, I think there is truth to the Indian, especially South Indian. I get small amounts of ASI and South Indian on various calculators (but so do most Europeans I suppose).

I think that logic makes sense. If Khoisan's extended all the up to East Africa, then perhaps some family members migrated, and other stayed? I would assume the migration route out of Africa would still be similar, it's more of a case of who went. If Oceanian populations have Indian mixing, it obviously suggests the path is correct. There was Austronesian migration back to Madagascar, and back to where they came from to a certain extent, which highlights that it isn't simply a case of just indigenous East African's leaving.

For me, the fact I get South African but not Sub-Saharan or Paleo African suggests it's specific to a population that descends from it. This is highlighted on the TreeMix images.

tamilgangster
10-20-2015, 03:56 AM
you have to take into consideration also, that even though they migrated from africa same can be said for all humans. Oceanian and africans have been seperated for 70000 years, and also the modern day populations in africa themselves are very different from the original OOA populations. These are ratios of less than one percents, so these numbers your getting are noise. Many OOA populations score bushmen or other african components on these runs at noise-ish levels at under 1 percent. So there is no real substance to support this claim.

BalkanKiwi
10-20-2015, 04:50 AM
you have to take into consideration also, that even though they migrated from africa same can be said for all humans. Oceanian and africans have been seperated for 70000 years, and also the modern day populations in africa themselves are very different from the original OOA populations. These are ratios of less than one percents, so these numbers your getting are noise. Many OOA populations score bushmen or other african components on these runs at noise-ish levels at under 1 percent. So there is no real substance to support this claim.

I disagree. There aren't studies to support it, however there are no clear studies that rule it out either. If it is noise, then it is clearly consistent among different data sets.

tamilgangster
10-20-2015, 06:42 AM
I disagree. There aren't studies to support it, however there are no clear studies that rule it out either. If it is noise, then it is clearly consistent among different data sets.

Your making a extreme conclusion off of very little evidence. There are no studies to rule it out because there hasn't been even an iota of legitamite evidence to support the claim. From the harappa DNA spreadsheet, the San component doesnt show up once for any of the Ocenian populations(tongan, samoan, papuan, aborigines) but it does show up sporadically at 1% among certain west asian and south asian populations(lebanese, kerala christian, baloch, pashtun, parsi, armenians, jatts, north ossetians, fiji indian, tajiks and turks). Its lacking among south arabians and north africans who are technically more african shifted. Its not found among east eurasian, european, or amerindian populations either

Awale
10-27-2015, 11:03 AM
This seems pretty unlikely but interesting, I suppose. The whole Australo-Melanesians and such are earlier migrants from Africa thing is mostly based on cranio-metric data & studies (http://eurogenes.blogspot.ae/2015/08/two-eurasians.html) which in this day and age are hardly relevant beyond maybe getting a semi-cool representation of what a pre-historic population might have morphologically looked like. From a genetic point-of-view once their Denisovan admixture is accounted for; the likes of Melanesians, Australian Aborigines and Papuans can clearly have Ust-Ishim modeled as "basal" to them the same way he is "Basal" to European Hunter-Gatherers & East Asians as well as Andamanese like the Onge (http://anthromadness.blogspot.ae/2015/09/is-eastern-non-african-situation-more.html). I even broached this subject with Iosif Lazaridis of Harvard Med:



Sheikh's message to him (http://oi58.tinypic.com/34sgzms.jpg)

His reply (http://oi62.tinypic.com/nxosih.jpg)



The dispersal Out-of-Africa is still poorly understood and it could be a lot more complex than just one jolt right out of the continent by one ancestral Eurasian population but one thing's clear from a genetic point of view... Their archaic admixture (Neanderthal + Denisovan) aside; all current data suggests that all out-of-Africa populations are essentially descended from one ancestral group (the Eurasian bottle-neck as Lazaridis puts it for example). So some model where a distinct group of pre-historic Africans departed and are ancestral to Oceanians but not to East Asians and Europeans simply doesn't cut it; they all seem to have diverged from one ancestral group and seemingly a while after the dispersal into Eurasia.

There's also the complicated fact that these populations tend to prove closest to the Andamanese & East Asians with various pop. geneticists like Lazaridis for now going as far as place them within a "clade" (Eastern Non-African) with these populations. Anyway, beyond that I'll only say that I find these populations having an actual "Southern African" element pretty unlikely.

puntDNALKing
11-21-2015, 06:17 AM
So, I think I solved this conundrum that vexed Jtoml4 and I. While I was reading a recent paper (click here (http://biorxiv.org/content/early/2015/10/18/029421)) that reconstructed the genetic history of Siberian and Northeastern European population, I saw a diagram that looked very similar to my treemix diagram (here (https://www.dropbox.com/s/hng1jbeylvihg9r/exampleone.jpg?dl=0) and here (https://www.dropbox.com/s/rl4s4cuobcamcid/exampletwo.jpg?dl=0)). However, this diagram included Neanderthal and Denisova Reference samples that changed the whole tree. The admixture between Khoisan and Oceania seen in my and dieneke's treemix was actually an admixture between a common ancestror of Denisova / Neanderthal and Oceania population. If you look at my and dieneke's treemix graph closely, you can see that the arrow is not directly on the khoisan end, but rather on the beginning of the branch. The reason our trees looked like that was because we were using khoisan as an outgroup population in order to create migration patterns.

Here is quote from the paper's data supplementary Figures and table.
"Analysis of Neanderthal-related admixtures among Oceanians. The full autosomal TreeMix model was calculated using a limited set of
eight individuals and two hominids. The most significant admixture was between a common ancestor of Denisova / Neanderthal and
Papuan. The amount of admixture was estimated at 5.6% with standard error of 0.58%. Standard error was obtained using a jacknife estimate."

The admixture I found in treemix was also around 6 to 7 %, so I think there is no connection between Khoisan (hg) and oceanian, but oceanian people are admixed with a common ancestor of both Neanderthal and Denisova.

parasar
11-24-2015, 04:26 AM
Your making a extreme conclusion off of very little evidence. There are no studies to rule it out because there hasn't been even an iota of legitamite evidence to support the claim. From the harappa DNA spreadsheet, the San component doesnt show up once for any of the Ocenian populations(tongan, samoan, papuan, aborigines) but it does show up sporadically at 1% among certain west asian and south asian populations(lebanese, kerala christian, baloch, pashtun, parsi, armenians, jatts, north ossetians, fiji indian, tajiks and turks). Its lacking among south arabians and north africans who are technically more african shifted. Its not found among east eurasian, european, or amerindian populations either

You have to remember Africans are very diverse now and they were very diverse at the time of OoA.
Imagine 1000 very diverse populations. Out of these 1000, one became OoA from which all non-Africans and many Africans descend. If not for the back-migration into Africa, even that 1% would be too high.

Now imagine that this one population first migrated to Oceania. While this looks unlikely, an alternative with the same effect is more likely - after Toba, the OoA got severely pruned and mainly (besides populations in Africa) those to the east of Toba survived. Effectively this would appear as if there had been a direct migration from Africa to Oceania.

BalkanKiwi
11-24-2015, 06:58 AM
You have to remember Africans are very diverse now and they were very diverse at the time of OoA.
Imagine 1000 very diverse populations. Out of these 1000, one became OoA from which all non-Africans and many Africans descend. If not for the back-migration into Africa, even that 1% would be too high.

Now imagine that this one population first migrated to Oceania. While this looks unlikely, an alternative with the same effect is more likely - after Toba, the OoA got severely pruned and mainly (besides populations in Africa) those to the east of Toba survived. Effectively this would appear as if there had been a direct migration from Africa to Oceania.

I like your thinking.

tamilgangster
11-24-2015, 10:28 AM
Judging by which populations score Khoisan on harappa DNA, and by the fact hes only .8% maori this still seems extremely unlikely. Its much more likely that the khoisan component is noise from the European portion of his ancestry.

ThirdTerm
02-22-2016, 03:57 AM
https://upload.wikimedia.org/wikipedia/commons/7/7a/HgE1b1b1a2.png

Haplogroup E-V13 from North Africa is especially common in the Balkans and it was found at particularly high frequencies in Kosovar Albanians (45.6%) and Macedonian Albanians (34.4%), and in some parts of Greece (35%). As a result, Europeans from the Balkans have some degrees of remote African ancestry but African haplogroups such as E (Y-DNA) and L (mtDNA) are almost completely absent in Oceania and there is no genetic evidence that sub-Saharan Africans directly migrated to Oceania.

Gravetto-Danubian
02-22-2016, 04:50 AM
https://upload.wikimedia.org/wikipedia/commons/7/7a/HgE1b1b1a2.png

Haplogroup E-V13 from North Africa is especially common in the Balkans and it was found at particularly high frequencies in Kosovar Albanians (45.6%) and Macedonian Albanians (34.4%), and in some parts of Greece (35%). As a result, Europeans from the Balkans have some degrees of remote African ancestry but African haplogroups such as E (Y-DNA) and L (mtDNA) are almost completely absent in Oceania and there is no genetic evidence that sub-Saharan Africans directly migrated to Oceania.

Well, everyone has some kind of remote African ancestry, but the E-V13 in southeastern Europe is from the Neolithic (probably), and came from E -M78 groups dwelling in the Levant rather than Africa directly.
In addition, E - haplogroups are northern African, not sub-Saharan (entirely different 'kind of people'). In fact, there is much debate going on as to whether haplogroup E represents a back-migration to northern Africa from Middle East.

parasar
02-23-2016, 03:51 PM
...

Haplogroup E-V13 from North Africa is especially common in the Balkans and it was found at particularly high frequencies in Kosovar Albanians (45.6%) and Macedonian Albanians (34.4%), and in some parts of Greece (35%). As a result, Europeans from the Balkans have some degrees of remote African ancestry but African haplogroups such as E (Y-DNA) and L (mtDNA) are almost completely absent in Oceania and there is no genetic evidence that sub-Saharan Africans directly migrated to Oceania.

There is almost no Y-E or mt L(xM, N) in South Asia either. Y-E was probably not born at the time of the exodus from Africa which I think was above the M168 level. Then most of the population of Eurasia died out (from YTT and related winter) leaving only pockets in SE Asia/Oceania east of Toba and in SS Africa. This distribution would make it appear that effectively there was a migration from South Africa to SE Asia/Oceania.

While Karmin et al. theorize based on M578 that a secondary expansion could have happened at the F level, I do not see any reason why it could not have happened at the CT and DE levels too.

Karmin et al.
"we observe an early split in our haplogroup F data that separates haplogroup G from HT-M578. However intersecting our data with Malaysian Chr Y sequence data (Wong et al. 2013) reveals a split in haplogroup F that predates the G/HT split by one mutation, F1329 (Figure S13). This finding is in accordance with the two Lahu F2-M427 individuals reported in Poznik et al. (2013) as having an ancestral allele of M578. In combination with the presence of deep branches of K in Southeast Asia, this further strengthens the model proposing that the initial radiation of the non-African Chr Y lineages may have taken place somewhere in Southeast Asia ... Only 24 mutational events distinguish the progression of two major non-African founder haplogroups F to K (Figure S13). Similarly small number of differences separate haplogroups LT, NO, S and P from their MRCA in haplogroup K (Figure S28), consistent with the suggestion of Karafet et al. 2014 (Karafet et al. 2014) that the initial diversification of Eurasian and Oceanian founder haplogroups was a rapid process limited to a few thousand years overall. We estimate that a peak of the coalescent events of the oldest non-African haplogroups falls into a time window of 47-52 kya"
http://genome.cshlp.org/content/suppl/2015/02/18/gr.186684.114.DC1/Supplemental_Text.pdf

StrandLoper
02-27-2016, 05:45 PM
Dienekes has posted this:

"Australia was one of the earliest regions outside Africa to be colonized by fully modern humans, with archaeological evidence for human presence by 47,000 years ago (47 kya) widely accepted [ 1, 2 ]. However, the extent of subsequent human entry before the European colonial age is less clear. The dingo reached Australia about 4 kya, indirectly implying human contact, which some have linked to changes in language and stone tool technology to suggest substantial cultural changes at the same time [ 3 ]. Genetic data of two kinds have been proposed to support gene flow from the Indian subcontinent to Australia at this time, as well: first, signs of South Asian admixture in Aboriginal Australian genomes have been reported on the basis of genome-wide SNP data [ 4 ]; and second, a Y chromosome lineage designated haplogroup C∗, present in both India and Australia, was estimated to have a most recent common ancestor around 5 kya and to have entered Australia from India [ 5 ]. Here, we sequence 13 Aboriginal Australian Y chromosomes to re-investigate their divergence times from Y chromosomes in other continents, including a comparison of Aboriginal Australian and South Asian haplogroup C chromosomes. We find divergence times dating back to ∼50 kya, thus excluding the Y chromosome as providing evidence for recent gene flow from India into Australia. "

Saetro
04-19-2016, 01:09 AM
Dienekes has posted this:

" The dingo reached Australia about 4 kya, indirectly implying human contact, which some have linked to changes in language and stone tool technology to suggest substantial cultural changes at the same time [ 3 ]. Genetic data of two kinds have been proposed to support gene flow from the Indian subcontinent to Australia at this time, as well: first, signs of South Asian admixture in Aboriginal Australian genomes have been reported on the basis of genome-wide SNP data [ 4 ]; and second, a Y chromosome lineage designated haplogroup C∗, present in both India and Australia, was estimated to have a most recent common ancestor around 5 kya and to have entered Australia from India [ 5 ]. Here, we sequence 13 Aboriginal Australian Y chromosomes to re-investigate their divergence times from Y chromosomes in other continents, including a comparison of Aboriginal Australian and South Asian haplogroup C chromosomes. We find divergence times dating back to ∼50 kya, thus excluding the Y chromosome as providing evidence for recent gene flow from India into Australia. "

This is from Bergstroem, Nagle, et al 2016 http://www.sciencedirect.com/science/article/pii/S0960982216000786
Looking closely into related publications, there is a slightly earlier publication from Nagle et al 2016, that seems to contain these plus other results. http://onlinelibrary.wiley.com/doi/10.1002/ajpa.22886/full

Nagle et al 2015 had to deal with the possibility that some submitters of test material might have had a European in their Y-line, so they discarded results from Haplogroups R, I, J, O as being non-indigenous.
"The single haplogroup H-M69 male in the sample knew that his great-grandfather was an immigrant from Europe and his Yfiler Y-STR profile indicated matches for his “minimum” haplotype of the YHRD with males from Europe, West Asia, and South Asia. As our study design only tested samples for haplogroup R-M207 in general (not its subhaplogroups), we cannot be certain that R-M479 (also known as R2), which is typical for South Asia, is absent from our sample (Table 1). However, all 125 haplogroup R 17-locus Y-STR haplotypes
that were not allocated to R-M17 or R-M412 after SNP typing were interrogated via the YHRD and could be assigned to either R-M420 or R-M343 with strong likelihood of European ancestry."

In summary, any samples that might have shown connections with South Asia were discarded, although further analysis of the R samples for possible R2 would have been useful.
These other samples are not even mentioned in the Bergstroem paper.
There are going to be some readers who feel their conclusion was due to circular logic - if you discard the only samples that could be recently linked, then there are no recent links - but this (apart from the R2 possibility above) is not supported.

But all of this is irrelevant to this thread: neither paper suggests any recent leaving of Southern Africa for Australia/Papua New Guinea.

BlessedbyHorus
06-26-2016, 02:14 PM
They are extremely distant from Africans according to genetic distance.