National Geographic Photo Books

Tuesday, October 31, 2006


If you are not already a member of the DNA-ANTHROGENEALOGY Yahoo! group, then I encourage you to check it out. It is very informative, with spirited discussion.

A recent thread involving the genetic makeup of Italy has provided some especially good reading. I don't think anyone will mind if I quote a little bit of it here:

I am going to speak from the perspective of a little remote mountain town called Montemurro in the Province of Potenza. This is in the Val d'Agri on the southern mainland. You do not get there by accident. It is the village of my grandfather's birth, but most importantly I hope to demonstrate the ethnic diversity found here in the example that if it is this diverse in such a remote and rural area, then imagine the implications for the coastal regions and larger cities and towns.

The village itself was rumored to have been founded by Saracens in the 9th century or so. It is definitely known that the Saracens destroyed the neighboring former Roman town of Grumentum in the 10th century and that the people of Grumentum relocated to Montemurro. Let me pause here for just a moment to explain the ethnic composition of Grumentum, which became the root stock of the village of Montemurro. Grumentum was founded in the 3rd century BC by the Romans. We might assume that the local population constituted at least some of the inhabitants. They were Lucani, also known as Lykoi from Anatolia. They had come to Italy in about 1300 BC and settled in the region of Lucania to which they later gave their name.

That bit was written by Dale J. J. Leppard, but the whole discussion is well worth reading.

Peer Review and SNP discovery

Yesterday, I encouraged folks who have a haplotype that resembles what I am calling "Eastern I1" and/or who have tested P38+ but are negative for downstream SNPs to engage Ethnoancestry to test EA's proprietary S23 and S31 SNPs. And I stand by that encouragement.

I failed, however, to mention that while these SNPs have been reviewed by ISOGG they have NOT been published or subjected to scientific peer review. Today, I was reminded of this fact by email.

Peer review is an important step in the scientific process, and anyone who undertakes a novel test like these S-series SNPs should be aware that they are ahead of the scientific community. The risk of being in this position is that these tests, which are not inexpensive, could be proven later to be a complete waste because they do not demonstrate what they purport to demonstrate.

As I said, I still hope that at least a few more folks have S23 and S31 tested but I want those folks to know that they are guinnea pigs. I don't mind being at the bleeding edge, and I know that many of the folks I correspond with don't either. It is important that we all remember that the bleeding edge is where we are.

Monday, October 30, 2006

Eastern I1* and R1b1b

One of the benefits of a project geographic project like this one is that occasionally we stumble across an interesting haplotype or two.

As I write this, we have just over 90 Italian y-chromosomes in the in Italy DNA Project. This is larger than many academic samples and our data is generally much more complete: all of our y-DNA participants have at least 12 STR markers and most have 25 or more. Nearly half the y-DNA participants have been genotyped for haplogroup.

Out of this data are emerging a couple of particularly novel haplotypes that we are studying.

One is a novel type of R1b1b, exhibited by kit numbers 39685 and 67866 (JSKGT and 7MARF at ySearch). These two folks have a distinctive profile of DYS447=22 and DYS449=32, which I have not observed in any other R1b participant in any public database.

Another is a group from Haplogroup I that test positive for M170, M258, P19, and P38 but negative for other I SNPs. Thus, according to the current ISOGG Y-SNP Tree, they are either I*, I1*, I1b*, or I1b2*. None of these are very common, and some are completely unobserved.

Unfortunately, Family Tree DNA does not offer advanced SNPs to further diagnose their haplogroup assignment so I am hopeful that they will engage Ethnoancestry to test the relevant S-series SNP tests and especially S31.

These "upstream" I haplotypes are an interest to Ken Nordvedt, I know, and seem to reflect a strong Eastern European origin. The group has a clear modal (ysearch ID 8V4MA) but enough diversity that STR-based "rules of thumb" are not yet apparent.

As we learn more about these and other haplotypes we will go into more detail here. Needless to say, however, that we are already starting to realize some of the benefits of a rapidly-growing geographical project and that network-effect benefits are starting to accrue.

Friday, October 27, 2006

Diagnostic Markers in Haplogroup G

We have a number of participants, some recently joined, that are in y-DNA haplogroup G or G2. I encourage these folks to check out an article by Phillip G. Goff and T. Whit Athey entitled "Diagnostic Y-STR Markers in Haplogroup G".

They found that four uncommonly tested markers (DYS425, DYS446, DYS452, and DYF399S1) can be successfully used to diagnose haplogroup G or its subgroups.

These are all available from Family Tree DNA as advanced tests, though DYS425 is part of the DYF371 test and DYF399S1 is part of the DYF399X test

Thursday, October 26, 2006

Are Italians underrepresented in DNA testing?

Are Americans of Italian descent underrepresented in genealogical DNA testing? I think the answer is yes.

To derive this conclusion, I examined ySearch, the largest public-access y-DNA database. As of today (26 October 2006), it contained a total of 33,570 records. Only 282 of those records identify someone from Italy as their most distant ancestor: less than 0.9%.

According to an analysis of the 2000 U.S. census, approximately 5.6% of Americans identified themselves as having Italian ancestry.

That is quite a gap, and can probably best be explained by an disproportionately high level of interest in genealogy on the part of folks from the British Isles.
EDIT: Ken Nordvedt pointed out that the Sorenson Molecular Genealogy Foundation y-chromosome database contains a much higher proportion of Italian samples: about 2.1%. It is speculated that the SMGF makes an effort to include a diverse array of ancestrys.
Of course I'd like to see more Italian-Americans undertake an interest in genealogy, but the main point is that those who DO take genealogical tests should not be surprised to find that their closest genetic matches are not Italian but English, Scottish, Irish, etc. Because those groups are overrepresented in the databases relative to Italians (or anyone of Mediterranean origin, actually), sheer chance is bound to produce a number of near matches.

Which DNA Test to Order?

One of the first decisions that someone new to genetic genealogy must make is which DNA test to order.

If you are primarily interested in your maternal line, then I have no problem recommending the basic mtDNA test from Family Tree DNA. It will provide a determination of your ancestral haplogroup, and a complete sequencing of the HVR1 region for a cost of $129. Even if you are lucky enough to find an exact HVR1 match, it will take a fair amount of serendipity to connect the matching family to your own. An HVR2 or full sequence upgrade is always available, and might be the first choice for those of unlimited means, but these tests are (in my opinion) best for the truly hardcore amateur geneticists.

If yo are primarily interested in your paternal line, the choice is a little tougher. However, I feel confident in strongly encouraging EVERYONE to order the 37-marker test right at the start. This test has the highest average mutation rate per marker and, and $189 per test, the lowest cost per expected mutation.

What I mean by this is that the combination of markers included in the 37-marker test is quite likely to produce at least one (and maybe two) differences between a pair of men who share a common ancestor from the early 1800s and to do so at the lowest possible cost. Since most Italian civil records are quite complete back to around 1810, conventional genealogical research should often be able to illuminate any relationship more recent than that. Since the purpose of DNA testing is to confirm a suspected relationship, you want a test with enough resolution to do so within about six or eight generations. The 37-marker test is the cheapest test that can do that.

And, if the 37 marker test reveals a close relationship, there are a couple of a la carte markers (DYS413, DYS445, DYS452, etc) than can be added to the basic panel for additional clarity for less than the cost of the 67-marker upgrade. Plus, the 37-marker test is quite robust when it comes to predicting haplogroups which is a valuable side-benefit.

Anyone wishing to join the Italy DNA Project can do so by following this link and ordering the Y-DNA37 test.

Sunday, October 22, 2006

100th Member

The Italy DNA Project is pleased to welcome it's 100th member!

We are excited that the project continues to grow, and that Italy's tremendous diversity is beginning to show through in the haplotypes of our members.

I'll remind you that you can always view the most recent project data at our FTDNA home page:

And anyone wishing to join the project can do so by following this link.

Monday, October 09, 2006

Advanced DNA Tests

Family Tree DNA has now made some new advanced DNA tests available to existing customers. Soon, new customers will be able to order them, too.

You will now see two options when you go to the "Order Tests" section on your personal page. The "Standard" option will offer the Y-DNA and mtDNA tests that FTDNA have previously offered. For all of typical genealogy group members the Standard section is the one for them. The "Advanced" option will offer these new tests which leads into three different areas.
  1. Expanded Y-STR offerings. These new offerings are y-DNA markers that differ from the markers included in FTDNA's standard panels.
  2. Autosomal STR offerings. Autosomal tests can be used, with some interpretation, to provide information about your ancestors' geographical location and in evaluating close family relationships (paternity, maternity, siblingship, etc.).
  3. X-chromosome STR offerings. Currently useful for close family testing, but possibly helpful in intermediate-term (a couple generations) testing as well.

I won't say much about the autosomal or X-STR tests, because at the moment those tests don't play a role in the Italy DNA project. And I should say that, although the new Y-STR tests are not particularly expensive (they are sold a la carte and in panels), if you have not yet upgraded to the 37 marker tests then you should do that before ordering any advanced tests. The 37 marker panel is a great value and is very helpful in helping answer both genalogical and anthropological questions.

Still, some folks in the Italy DNA Project might be interested in some of the new Y-STR markers. They can help clarify your relation to other folks who are a close match at 37 markers and they can sometimes provide insight into your deep ancestry.

Folks in haplogroup E3b might consider looking at DYS413. This marker is highly mutative, and can be used to segregate E3b into some of its subgroups. Folks in haplogroup J2 might consider both DYS413 and DYS446 for the same reason. Folks in R1b might look at DYS413, DYS463, DYS464x, and Y-GATA-A10.

Also, some markers offered by Sorenson labs (e.g. SMGF, Relative Genetics, and DNA Heritage) are now offered as part of these new advanced tests. Ordering these markers can bring your FTDNA results into full compatibility with relatives who tested at these other labs. The markers that accomplish this are: DYS441, DYS444, DYS445, DYS446, DYS461, DYS462, DYS463, DYS635 (aka Y-GATA-C4), Y-GATA-A10, & Y-GGAAT-1BO7.