No studies necessary: do your own replication!

By Razib Khan | December 16, 2011 9:22 pm

In response to the post below I received the above response on twitter. This is an interesting case. The link goes to a paper in the year 2000, Alu insertion polymorphisms in NW Africa and the Iberian Peninsula: evidence for a strong genetic boundary through the Gibraltar Straits:

An analysis of 11 Alu insertion polymorphisms (ACE, TPA25, PV92, APO, FXIIIB, D1, A25, B65, HS2.43, HS3.23, and HS4.65) has been performed in several NW African (Northern, Western, and Southeastern Moroccans; Saharawi; Algerians; Tunisians) and Iberian (Basques, Catalans, and Andalusians) populations. Genetic distances and principal component analyses show a clear differentiation of NW African and Iberian groups of samples, suggesting a strong genetic barrier matching the geographical Mediterranean Sea barrier. The restriction to gene flow may be attributed to the navigational hazards across the Straits, but cultural factors must also have played a role. Some degree of gene flow from sub-Saharan Africa can be detected in the southern part of North Africa and in Saharawi and Southeastern Moroccans, as a result of a continuous gene flow across the Sahara desert that has created a south-north cline of sub-Saharan Africa influence in North Africa. Iberian samples show a substantial degree of homogeneity and fall within the cluster of European-based genetic diversity.


There are two issues at work here. A minor one is that I don’t necessarily disagree with the above study. The dense SNP-chip analyses do indicate a major division at the Mediterranean, and a preponderance of the ancestry of modern Iberians pre-dates the Moorish period. My only contention is that on the order of ~10 percent does seem to possibly derive from an influx of Berbers and Arabs, and this is far higher proportionally than the current fashion among historians, who tend to hold that migrations in the post-Roman world were of very small numbers, if they were demographic at all. My own suspicion is that what we’re seeing is partly reproductive variance by class. A small number of elite lineages may have a large demographic impact over the long term.

But there’s a bigger point, and that’s that of open science and what “counts” as evidence in a given argument. Frankly, an 11 marker study from the year 2000 holds less sway for me than analyses of hundreds of thousands of SNPs today. Just because Dienekes doesn’t publish in journals doesn’t mean that it’s not worth citing seriously. In fact, some of the posts that he puts out are more meaty than what you might find in some journals! But there’s a bigger issue: I’ve seen the same patterns as Dienekes in many cases, and that’s why I believe his results have some validity. In other words, I’ve replicated some of his findings. So has Zack and David. Sometimes there is disagreement. For many analyses and inspections of population structure you don’t need to look at the academic literature. That’s why I’ve put up a simple tutorial with scripts, how to use them, and a modest data set to go along. The academics are essential for proposing more powerful analytic techniques, subtle interpretations, as well as obtaining data sets (e.g., ancient DNA) which others would not be able to get a hold of. But the number of data sets in the public domain are so numerous one shouldn’t be citing something from even a few years ago at this point. The History and Geography of Human Genes is a great book, but you have more ability to crunch and analyze the underlying populations today in one day than the author of that book had for decades before its being published!

MORE ABOUT: Human Genomics
ADVERTISEMENT

Comments (8)

  1. “North African” haploid ancestry in Iberia follows a West-East cline and therefore does not correspond at all with the Muslim period but with an older founder effect, maybe in the Neolithic or maybe even as old as the Solutrean/Oranian probable interaction (which would also explain the 25% of Iberian-derived mtDNA and maybe 10% Y-DNA in North Africa).

  2. Karl Zimmerman

    Okay, noting my original response was dealt with in the comments to the other post.

    Seems an easy way to decipher if the Northwest African/Southwest Asian component is the result of settlement of Iberia (Phoenician, Carthaginian, Moorish), or noise due to the zombie constructs, is to include all French populations in the analysis. If French populations also show elevated levels, it suggests there was perhaps a middle-eastern component to the Indo-European migrants, or that a non-IE population introduced farming prior to the IE expansion.

  3. or noise due to the zombie constructs, is to include all French populations in the analysis. If French populations also show elevated levels, it suggests there was perhaps a middle-eastern component to the Indo-European migrants, or that a non-IE population introduced farming prior to the IE expansion.

    i thought i made it clear it clear that the french did?

    “North African” haploid ancestry in Iberia follows a West-East cline and therefore does not correspond at all with the Muslim period but with an older founder effect, maybe in the Neolithic or maybe even as old as the Solutrean/Oranian probable interaction (which would also explain the 25% of Iberian-derived mtDNA and maybe 10% Y-DNA in North Africa).

    can you give me a citation to dig into? i’m more suspicious of haploid ancestry. e.g., the indian mtDNA is very unrepresentative of total genome relationships, probably due to ancient sex-biased migration.

  4. Phil

    “or noise due to the zombie constructs, is to include all French populations in the analysis. If French populations also show elevated levels, it suggests there was perhaps a middle-eastern component to the Indo-European migrants, or that a non-IE population introduced farming prior to the IE expansion.

    i thought i made it clear it clear that the french did?”

    Then your conclusive statement in the previous post (that regarding moorish ancestry) is irrelevant and then Maju’s last comment turns to be true.

  5. #4, i meant the french show the west asian component. they don’t show the north african/southwest asian one. didn’t #2’s comment close enough. the latter is a hallmark of iberians in european samples, with the exception of sicilians and such.

  6. also, with all do respect, people need to shut the fuck up and start running their own analyses if they want to talk about this. there’s a lot of stuff to argue about, but it’s hard to communicate it with people who don’t have a first hand sense of the range of outcomes.

  7. “can you give me a citation to dig into?” Adams 2008 (http://www.cell.com/AJHG/fulltext/S0002-9297%2808%2900592-2) for Y-DNA. I do not have such a clear reference right now for mtDNA (notably U6) but I have the foggy impression that it is the case as well.

    Other references of interest on the North African-SW European connections (my blog posts, where you can find the academic reference):

    http://leherensuge.blogspot.com/2010/05/new-paper-on-mediterranean-genetics.html (Tolouse is closer to North Africa than Andalusia, go figure!)

    http://leherensuge.blogspot.com/2010/08/genetic-flow-across-strait-of-gibraltar.html (interesting comparison of various genetic markers across the Strait of Gibraltar)

    Also a reader and occasional collaborator (I consider him quite well informed and with interesting opinions) analyzed the Iberian West-East cline (Y-DNA): http://forwhattheywereweare.blogspot.com/2011/04/guest-post-by-argiedude-west-east-y-dna.html

    The essence of the matter is that there is much more “North African” Y-DNA (E-M81 essentially, J1 is very rare in Iberia overall and may have arrived later and separately) in NW Iberia, which was the core of the “Reconquista” Christian realms (Asturias-León, including Galicia & Old Portugal) than in SE Iberia where the “Moors” persisted for the longest (eight centuries or so). However J1 may correlate with the Muslim intrusion but J1 is never more than 3% locally (SE Iberia, South Portugal, Valencia), while in North Africa it is 8-28% (and it can also have come earlier with the Roman Empire, Phoenicians, Megalithism…)

    So IMO J1 in Iberia correlates with “recent” North African influence at quite low levels, while E-M81 reflects instead a much older founder effect, which can be Neolithic or maybe even older (its strong presence in Asturias and Cantabria but not among Basques is quite hard to explain).

    “people need to shut the fuck up and start running their own analyses if they want to talk about this”.

    You do have a point and the merit of having explained in the past how to do it: http://blogs.discovermagazine.com/gnxp/2011/03/analyzing-ancestry-with-admixture-step-by-step/

    I’m, as we speak, downloading and installing the programs (luckily I use Linux). I hope to figure out all the process some day.

  8. Regarding to mtDNA (U6 notably) the open access reference is http://www.ncbi.nlm.nih.gov/pmc/articles/PMC1852723/ (Achilli 2007), which indeed detects more U6 towards the West, specially in North Portugal (8%) and Central Portugal (5%), while the South only has 1-2%. (other regions of Iberia often 0%, excepting “Spain Central”, which has 6% and is a Maragato sample, from a mountain area near North Portugal).

    So the tendency in both Y-DNA E-M81 and mtDNA U6 seems to be a West to East cline and even with a tendency to concentrate the “North African” lineages around North Portugal.

    In the discussion on Argiedude’s analysis, there is a lot of information that was only vaguely known to me, for example Casas 2006 (PPV) found lots of L(xM,N) in an medieval post-reconquista but pre-Atlantic slave trade (13th century) site near Córdoba – while today they have some of that, it is only a fraction. Also again the Maragatos from NW Spain (near León) show high apportion of L(xM,N). I contend that it may have come from North Africa together with Y-DNA E-M81 and mtDNA U6.

NEW ON DISCOVER
OPEN
CITIZEN SCIENCE
ADVERTISEMENT

Discover's Newsletter

Sign up to get the latest science news delivered weekly right to your inbox!

Gene Expression

This blog is about evolution, genetics, genomics and their interstices. Please beware that comments are aggressively moderated. Uncivil or churlish comments will likely get you banned immediately, so make any contribution count!

About Razib Khan

I have degrees in biology and biochemistry, a passion for genetics, history, and philosophy, and shrimp is my favorite food. In relation to nationality I'm a American Northwesterner, in politics I'm a reactionary, and as for religion I have none (I'm an atheist). If you want to know more, see the links at http://www.razib.com

ADVERTISEMENT

See More

ADVERTISEMENT

RSS Razib’s Pinboard

Edifying books

Collapse bottom bar
+