Your Dr. McDonald Results

[Scarlet Ibis]

Project Description

Dr. Doug McDonald's Biogeographical Ancestry (BGA) Project is focused on auDNA and has been quite popular in the genealogy world for a good few years. It was one of the most informative means of exploring one's recent ancestry before the "Genome Blogger" era started roughly three years ago. As with the others, his aim is to shed light on the genetic structure of world populations while simultaneously providing a free informative service to enthusiasts.

Dr. McDonald receives data from personal genetics customers (zipped format) who have tested with 23andMe and runs their results through his "in-house" software. He uses the HGDP populations and his own private database as references.

Participants in his project receive a chromosome painting with ADMIXTURE-like component breakdowns, charts displaying their genetic position and numerical data showing their "proportional fit" with his references.

Please contact an administrator for Dr. McDonald's email address.

---

Mine:

Most likely fit is 47.9% (+- 0.0%) Europe (all Western Europe)
and 52.1% (+- 0.0%) E. Asia (all Mongolian/Siberian)

The following are possible population sets and their fractions,
most likely at the top
Irish= 0.479 Mongolia= 0.521

I would not take those exact populations as really true, but
it very clearly is 50-50 far western Europe and northern East Asia
but not really Siberia.

The American on the chromosomes is probably not real.

The overall fit quality is abysmal, and I could not do better by hand.
This likely means that one or the other component is some unusual
group I do not have as comparison.

Doug McDonald

My Mom's:





My Dad's:

[DavidCar]

I'm trying to understand more about Dr. McDonald's project, as I have some 23andme data that could be analyzed. A Google search sends me to an ISOGG page which only contains links to two now inaccessible discussion threads on DNA Forums. (http://www.isogg.org/wiki/McDonald%27s_BGA_project). It might be useful if another description of the project was posted here.

[DMXX]

DavidCar,

Please re-check the first post - Scarlet Ibis has inserted a quick description of McDonald's BGA and participation.

On the new test you appear as almost pure Iranian, with just a little
(3-10%) of something more northerly ... anything from Ireland to Siberia
fits reasonably well. This means that you are just a bit from
more northerly regions than central Iran, if unmixed. Of course,
this is not exact ... the average error is a couple of hundred miles
inside Europe; in the Mideast I'd expect it to be Iran or
very close to it.

[Artemis]

Hmm I never received the chromosome painting but then again the last time I sent anything in for testing was a very long time ago. (June 2011 to be exact). Should I resend?

[DMXX]

Hmm I never received the chromosome painting but then again the last time I sent anything in for testing was a very long time ago. (June 2011 to be exact). Should I resend?

I would say it's a good idea. As elaborated here, McD's Painting gives you a decent "low-fi" place to start off from when identifying segments which deviate from the main contingent of your genetic make-up. In my case, it's the "East Asian" segments.

Dr. McDonald's BGA can be treated as a K=6 ADMIXTURE run; one step up from 23andMe's Ancestry Painting, but not as specific as the genome blogger efforts.

[Seabass]

The automatic program is not doing that great, saying you
are 50% Jewish, which does not fit the special Jewish test.

So I did special tests and get

French 0.6322 Jewish 0.0922 Armenian 0.2756 or
Tuscan 0.2735 Romania 0.6421 Jewish 0.0844 or
Italian 0.2005 Romania 0.6966 Jewish 0.1028 or
Tuscan 0.4831 Hungary 0.3716 Armenian 0.1453 or
Spain 0.4001 Hungary 0.3146 Armenian 0.2853 or
Tuscan 0.2445 Romania 0.6829 Armenian 0.0725 or
French 0.6147 Jewish 0.1098 Georgian 0.1275 Cypriot 0.1479

I’m just not sure ... it clearly is some sort of Mideast mix with France or Italy or
possibly Spain. But the actual amount of real Jewish is in doubt,
as is the presence of something from the Caucasus. There could be
something like say Serbian.

Doug McDonald

LOL I'm doing his head in! I fear his tests could prove to be very bad for say an adopted person of mixed ancestry, though fortunately I'm not in this boat and can make sense of these results. Its a good idea to follow some gedmatch oracle results before sending him your data. While I think his scripted program is unique in its own merit and has a lot of potential, I think he is cutting himself short a little by not updating his population sample database.

[Spark]

LOL I'm doing his head in! I fear his tests could prove to be very bad for say an adopted person of mixed ancestry, though fortunately I'm not in this boat and can make sense of these results. Its a good idea to follow some gedmatch oracle results before sending him your data. While I think his scripted program is unique in its own merit and has a lot of potential, I think he is cutting himself short a little by not updating his population sample database.

Seabass, what is your actual ancestry and how did McDonald's analysis deviate from what you would have expected? I know he's not perfect and some ancestries and ethnicities are dicey, but he tries.

[Seabass]

Seabass, what is your actual ancestry and how did McDonald's analysis deviate from what you would have expected? I know he's not perfect and some ancestries and ethnicities are dicey, but he tries.

Don't get me wrong, I do appreciate and enjoy his analyses along with how quick his response is. To be honest I did see these results coming, though if he had more reference populations the results would not be genealogically inaccurate as they are. Here is what he gave me a year ago. These are probably the same results he got that he wasn't happy with hence he tried some 'special tests' since he did not think I could be split in two using his reference populations.

Most likely fit is 59.2% (+- 9.5%) Europe (all Western Europe)
and 40.8% (+- 9.5%) Mideast (various subcontinents)

The following are possible population sets and their fractions,
most likely at the top
English= 0.421 Jewish= 0.579
French= 0.654 Armenian= 0.346
French= 0.692 Druze= 0.308
French= 0.514 Jewish= 0.486
English= 0.608 Druze= 0.392
Spain= 0.662 Georgian= 0.338

However, a look at the PCA "scatter" plot with the Jews at the top
says you are not half-Jewish (which would have you half
way between the Jews and England/France) so the other
possibilites are more likely.

Doug McDonald

[Seabass]

2zpvbib.jpg

qy758w.png

35c1uyv.png

2isjeqa.jpg

eqpjiq.jpg

[apophis99942]

Most likely fit is 45.3% (+- 2.3%) Africa (all West African)
and 18.2% (+- 1.5%) Africa (all East African)
which is 63.6% total Africa
and 26.2% (+- 1.1%) Europe (all Northeast Europe)
and 10.3% (+- 0.5%) E. Asia (various subcontinents)

The following are possible population sets and their fractions,
most likely at the top
Mandenka= 0.435 Maasai= 0.195 Russian= 0.273 Japanese= 0.097 or
Mandenka= 0.436 Maasai= 0.193 Russian= 0.266 Tu= 0.105 or
Mandenka= 0.434 Maasai= 0.196 Russian= 0.271 North Ha= 0.098 or
Mandenka= 0.489 O-Ethiop= 0.161 Finland= 0.249 Japanese= 0.102 or
Mandenka= 0.489 O-Ethiop= 0.159 Finland= 0.241 Tu= 0.110 or
Mandenka= 0.443 Maasai= 0.186 Finland= 0.261 Tu= 0.111 or
Mandenka= 0.442 Maasai= 0.188 Finland= 0.268 Japanese= 0.102 or
Mandenka= 0.436 Maasai= 0.194 Russian= 0.275 South Ha= 0.094 or
Mandenka= 0.488 O-Ethiop= 0.162 Finland= 0.247 North Ha= 0.103 or
Mandenka= 0.441 Maasai= 0.189 Finland= 0.266 North Ha= 0.104