The GEDmatch MDLP Project (Explained For Beginners)

The GEDmatch MDLP project has a mix of calculators that compare your ethnicity with DNA samples assembled by the project creator.

Like the other calculators, this project measures how your DNA aligns with ancient samples alongside results from project volunteers and academic studies. The focus is on exploring your heritage from older eras of human history.

This article gives a detailed review of the most popular MDLP calculator, the World-22, which is most suited to European and Eurasian heritage. We also have tips for the other MDLP calculators.

Origins Of The MDLP Project

The creator of the MDLP project is an Estonian blogger named Vadim Verenich.

In 2011, Verenich started a new project to focus on the heritage of an area covering the former Grand Duchy of Lithuania. This is also known as Magnus Ducatus Lithuaniae. He started a blog and issued a call for volunteers to send him DNA samples if their grandparents were:

  • Belarusians
  • Lithuanians
  • Poles (eastern)
  • Latvians (from Latgale)

Verenich was helped get his project off the ground by Davidski of Eurogenes and Zack of HarappaWorld. Like the others, he modified the admixture software from Dienekes of Dodecad and created his own set of calculators for the MDLP project.

The GEDmatch MDLP project dates back to 2012-2013. Bear in mind that all these GEDmatch projects are quite outdated. There have been recent discoveries of archaic DNA that are not reflected in MDLP. And more people have tested their DNA from diverse regions that aren’t represented within the projects.

However, the calculators can be very interesting to explore.

Expanding The Populations

Verenich expanded the geographic reach of his project through 2012. Having diverse populations as a reference can only help isolate and analyze the original targets

MDLP Populations

You’ll see that most of the MDLP calculators have a number within their name: 11, 16, 23, and 22. You can take it from me that the one that doesn’t (“World”) should have the number 12.

These are the K numbers. It’s the count of populations or communities that are analyzed by the calculator. The “k11 modern” uses eleven populations, while “world-22” uses 22 populations.

Your results show percentages across these populations. Don’t assume that a higher number of populations makes for better results. Unfortunately, the opposite is true. The levels of error increase as the K number rises. So, the more detailed calculators are also more speculative.

Beware The Best Fit

Unfortunately, just because a calculator has fewer populations…doesn’t make it better for you.

A problem with all calculators is that they will do their best to fit you into the model.

Let’s say your DNA is predominantly Irish. There are very few Irish or Scottish samples in this project. Your results from the K16 calculator are likely to drop into the North East Europe bucket because it’s the nearest alternative.

GEDmatch MDLP World-22 Calculator

The MDLP World-22 calculator uses twenty-two reference populations when analyzing your admixture.

The original project focused on Belarus, Lithuania, Eastern Poland, and the Latgale region of Latvia. Ultimately, Verenich collected about 90 samples from volunteers.

The project expanded through cherry-picking DNA samples from various academic genome studies. These include:

  • Bulgarian, Ukrainian, and Mordovian samples from the Yunusbayev study
  • Romanians, Uzbeks, Turks, Armenians from Behar
  • Russians and Adygei from HGDP
  • British and Finns from the 1000 Genomes Project
  • Neolithic DNA samples found in Sweden
  • Mesolithic DNA samples found in the Iberian peninsula

Verenich also grabbed ten samples from each European country represented in the POPRES study.

Understanding The Populations

There are bound to be some populations amongst the list of 22 that aren’t clear in their meaning. How about Atlantic-Mediterranean-Neolithic? East-Siberian? What do these mean?

The Oracle spreadsheet lets you review the groups of DNA samples that make up each of these populations. We have a tutorial that walks step by step through using the Oracle spreadsheet.

Although the tutorial uses the Dodecad project as an example, the steps apply to all the spreadsheets. But I’ll walk through an MDLP example here.

Suppose I want to understand more about the East-Siberian population. It’s the 13th population listed across the top row of the MDLP-22 spreadsheet. You then run your eye down the groups on the left, looking for the higher percentages.

The display is awkward to use like this. Our tutorial shows you how to copy it to a proper spreadsheet, like Excel. When I get the reference data into Excel, I can put a filter on the column to show groups (rows) with a percentage of above 50.

The shortlist includes Evenk (a people of the Nothern Asia) and Yakut (a Turkic ethnic group).

Another group called “Even” illustrates a recurring problem across the projects. I’m sure that this is another Evenk group. Similar-sounding groups often crop up.

The fourth group is named “East-Siberian”, which is the same as the broad category. I think this is what’s known as a synthetic population, one that is derived by the project creator through statistical methods.

Comparing Your Breakdown To The Spreadsheet

Let’s turn our attention back to the Evenk and Yakut groups. The reference spreadsheet shows you their admixture breakdown across the 22 populations.

If the percentages of your own admixture are closely aligned to their breakdown, this becomes an interesting insight. Just bear in mind that we’re dealing with pre-modern eras.

Interpreting Your World-22 Oracle Results

The next step is to use the Oracle utilities. The MDLP projects have both the Oracle and Oracle-4 utilities.

These may be useful when your grandparents are from different regions. The statistical display isn’t obvious to interpret. We have an article with a full walkthrough of the GEDmatch Oracle.

It goes through detailed examples using the Dodecad project, but the principles are the same.

Learning From Other Peoples’ Results

It can be useful to see other people’s results that are similar to your own, particularly when they describe their heritage. You can check out this thread on the Anthrogenica forum where people post their World-22 results.

Some posters just give their results with no background. But if you trawl through the pages, you’ll find examples where people give a rundown of their known ancestry.

You’ll also see some interesting comments from people who have spent time delving into all the GEDmatch projects. This is one example:

This is probably the first calculator I’ve run into which does a good job of separating out South Asian accurately…it actually pulled out my Oceanian/East Asian suitably. Probably because there’s a ton of East Asian components here, though I’m not sure how accurate they all might be.

Dr. McNinja, Anthrogenica

MDLP World Calculator

The World calculator only has twelve populations. Here’s a thread from Theapricity forum on reactions to this calculator.

The consensus is that the smaller number of populations are skewed eastwards. So, Western and North-west Europeans get too much representation from eastern European regions.

And other commentators report that Serbs and Croats are shifted southward.

I suggest you stick with the World-22 option.

GEDmatch MDLP K11 Modern

I have no idea why “Modern” is tacked onto the end of the K11 calculator. This calculator is more concerned with mesolithic and neolithic periods.

I suppose that the neolithic era would be considered “modern” if you were a disapproving individual at the tail-end of the mesolithic!

One problem with this calculator is that the Oracle utility doesn’t work. You’ll get an error message if you try to use it. This hasn’t been fixed in nearly ten years, so I’m not expecting it to work anytime soon!

Don’t be confused by the abbreviations used for some populations. Here’s a rundown…

  • EHG: Eastern Hunter Gatherer
  • WHG: Western Hunter Gatherer
  • ASI: Ancient South Indian
  • SEA: South East Asian

If you’re interested in the Hunter Gatherer side of things, then you may want to compare these results with the Eurogenes Farmers vs Hunter Gatherers calculator. In the linked article section, I give the (very) basics of the anthropological background.

If you want to see the results of other people using this calculator, check out this Anthrogenica thread.

GEDmatch MDLP K16 Modern

This calculator is a little more modern than the K11, although it still has plenty of ancient DNA samples. And the Oracle utility works with this calculator.

The oddly named “Ancestor” population is African.

The “Steppe” population caused a degree of confusion in the anthropology forums. This is a recurring problem across the GEDmatch projects, where the creators co-opt a term that has a different academic interpretation.

You can see the K16 results from other people on these forum threads:

In general, there were mixed reactions to this calculator. A commentator of West European descent labeled his results as “incoherent”, while an Indian poster said it was “pretty good”.

MDLP K23b Calculator

The MDLP K23b calculator uses twenty-three reference populations.

It’s actually quite different to the World-22, which of course has one less population. While K23b has six African populations versus three in World-22, the K23b has one indigenous American population versus five in World-22.

If your heritage is South Asian, I’ll direct you to a lengthy comment by Dr. McNinja on this page of an Anthrogenica thread. The interesting analysis may be relevant to your own research.

The Anthrogenica thread is also interesting because Vadim Verenich, the MDLP project creator, is answering questions in the early pages. You can see him start chiming in on page 2.

MDLP And The Calculator Effect

I mentioned in the previous section that the project creator was answering questions for a while on the Anthrogenica forum. This is what he has to say about inaccurate calculator results.

By the way, one can’t blame “the calculator effect” for producing the strange results. There is no calculator effect in the K23b, because I created this calculator according to Polako’s [Eurogenes] guidelines.

Vadim Verenich, Anthrogenica

Polako was a nickname for the creator of the Eurogenes project. By “the calculator effect”, Vadim is referring to a potentially fatal flaw that the Eurogenes creator said was in the original Dodecad calculator and methodology.

All the GEDmatch calculators are based on the Dodecad software and methodology.

The “Calculator Effect” would mean that the admixture results were only good for the original volunteer samples included in each project. Every other user (that’s you and me) would get low-quality results.

But Davidksi (aka Polakao aka Eurogenes) made changes to the methodology that he claimed would fix the flaw. And he wrote several blog posts about it.

The Dodecad project creator, Dienekes, fiercely disputed these claims. He basically said that his version was right and that the Eurogenes amendments introduced a fatal flaw to results.

This descended into an online flame war with silly name-calling. If you’re interested, here’s Davidski’s original argument, and Dienekes’ first rebuttal.

I like to sum it up with double-spiderman.

spiderman pointing at spiderman

I think the main take-away is that all the calculators are fascinating, but you shouldn’t take the results too seriously.

Alternative GEDmatch Projects

Check out our detailed articles on other admixture projects:

Researching Modern Ethnicity

The GEDmatch projects aren’t aimed at the modern era. The ethnicity breakdowns provided by the consumer DNA companies are oriented toward more recent generations. However, their results can also be very questionable.

Both Ancestry and MyHeritage have a more detailed breakdown of ethnicity that’s based on your DNA matches on those sites. Both have been very accurate for my Irish heritage.

Our article on interpreting Ancestry’s ethnicity estimates has a detailed section on their genetic communities. And we have a separate article on MyHeritage genetic groups.

More Articles And Tutorials

.

Margaret O'Brien
Latest posts by Margaret O'Brien (see all)