Overestimated Admixture in Brisighelli (2012)

September 8, 2014

This study tries to quantify individual ancestry in Italians by using a small panel of autosomal markers that are known to produce errors and overestimate admixture compared with using the full SNP data set, and as a result it comes up with estimates of Sub-Saharan African admixture in Italy and elsewhere in Europe that are much higher than in any other study:

A panel of 52 AIMs was genotyped in 435 Italian individuals in order to estimate the proportion of ancestry from a three-way differentiation: sub-Saharan Africa, Europe and Asia. Structure analyses allowed us to infer membership proportions in population samples, and these proportions can be graphically displayed, as in Figure 2. This analysis indicated that Italians have a basal proportion of sub-Saharan ancestry that is higher (9.2%, on average) than other central or northern European populations (1.5%, on average). The amount of African ancestry in Italians is however more comparable to (but slightly higher than) the average in other Mediterranean countries (7.1%).

The authors go on to say that Sub-Saharan African admixture in Italy is also evident in uniparental markers, but they should've been suspicious of their high AIM-based estimate when they found only 1.2% mtDNA haplogroup L and no A, B or E(xE3b) Y-DNA, which leads to an admixture estimate (0.6%) that's 15 times lower.

They also should've been puzzled by other bizarre results in their structure analysis, like African admixture appearing higher in Northern and Central Italy than in Southern Italy, with an equally high amount in Britain and Japan, and a very high Asian component in Central and Southern Italians that other Europeans don't have:

52 AIMs

Indeed, when we look at many of the same and similar populations tested using 291,184 SNPs in Lazaridis et al. (2014), we can see that the pattern is totally different and fits much better with the uniparental data and known reality. Italians only have a drop (<1%) of African admixture, which is highest in the south, while the British and Japanese don't have any at all, and Asian admixture is expectedly higher in Northern and Eastern Europe than it is in Southern Europe:

291,184 SNPs

These two very different results highlight the importance of using all genome-wide data when estimating individual ancestry and admixture proportions.

---------------
Brisighelli et al. "Uniparental Markers of Contemporary Italian Population Reveals Details on Its Pre-Roman Heritage". PLoS One, 2012.

2 comments

Crimson Guard said...

While levels are negligible, they're both crap done up on computer programs, sometimes of authors own design.

That Hellenthal et al. found the high proportions of SSA admixture in Lithuanians and Northern Europeans but the idiots chose to ignore it as "noise" while only reported it on Southern Europeans.

http://s1.zetaboards.com/anthroscape/topic/5383284/1/

With the studies you posted, they're probably picking up noise that they've added up or something they interpret as non-European(as in Northern European) as "North African-like" or markers for bloodtype or sickle cell and turning it into SSA.

Sometimes Bergamo shows MENA admixture under certain authors other times it doesn't. In fact most of these studies yield different results and not sure how reliable they are as most geneticists are Socialists and can be lumped into the race denial, multicultrist, Orientalist, Afrocentrist or Nordicist incarnations of anthropology as pretty much that is what you find in schools.

Probably also depends on what methods and/or their interpretation of their programs. Or simply some scientists are biased agenda driven.

Lazaridis et al. (2014) looks to be making the same errors as Moorjani did, and he is likely referencing him.
Besides his sample sizes are ridiculously small.

They're probably picking up E3b and mtDNA M1 and mis using "African", sometimes they have Australoid showing up Europeans,lol. Also if Northern Europeans have a Mesolithic ancestry than they too should have indirect SSA or Negritoid admixture plus whatever East Asian.

Looks like they're probably playing with their programs and inflating SSA admixture in Middle Eastern populations to get more visibility of Asian admixture in Northern Europeans so intermediate Euro-mediterranean peoples end up as intermediate between this East Asian shifted (really continental mainland and Eastern) Europe and the Middle East.

Moreover, Di Gaetano et al didnt find any sub-Saharan admixture in Italy. While even Utah(Northern Europeans) whites had slightly higher North African admixture:

http://italianthro.blogspot.com/2012/10/population-structure-within-italy.html

I think this new study is the clincher, so we can move past retarded reported levels of North African/MENA admixture in Southern Italy/Sicily and bogus east west divide of the island.

"Besides the low frequency (1.5%) of E-M81 lineages in general observed in our SSI dataset, the typical Maghrebin core haplotype 13-14-30-24-9-11-13 [8] has been found in only two out of the five E-M81 individuals. These results, along with the negligible contribution from North-African populations revealed by the admixture-like plot analysis, suggest only a marginal impact of trans-Mediterranean gene flows on the current SSI genetic pool. Together with the Berber E-M81, the occurrence of the Near-Eastern J1-M267 in Southern-European populations has been linked to population movements from the Near East through North-Africa, and particularly as a marker of the Islamic expansion over Southern-Europe (started approximately in the 8th century AD and lasted for more than 500 years). Fisher exact tests based on HGs frequencies have revealed the presence of haplogroup J1-M267 at significantly higher frequencies in both North-Africa and the Levant than in Sicily and Southern Italy (both P-values<0.001). However, the estimated age for Sicilian and Southern-Italian J1 haplotypes refers to the end of the Bronze Age (3261±1345 YBP), thus suggesting more ancient contributions from the East. Nevertheless, our time estimate does not necessarily coincide with the time of arrival of J1 in SSI; in fact a pre-existing differentiation could potentially backdate the time estimate here obtained."

http://www.plosone.org/article/info%3Adoi%2F10.1371%2Fjournal.pone.0096074

Italianthro said...

They're not both crap. Only the first one with the AIMs. The difference is obvious.

What you're talking about in Hellenthal et al. is not an admixture analysis. It's a new method called Globetrotter that seems to have some limitations. But their Admixture analysis shows similar results to Lazaridis.

Moorjani et al. is also using a different method called PCA Projection that has problems. It's got nothing to do with Lazaridis' admixture analysis.

Di Gaetano et al. didn't include any Sub-Saharan African samples in their study, so you can't say they didn't find admixture that wasn't even tested for.

That last study is fine, but it's on North African admixture and only looks at uniparental markers, so no autosomal admixture analysis or SSAs.