New Pop Finder?

[Stevo]

Looks Good

I had a look at the new "myOrigins" thingy this morning, and I like it. IMHO, it is a BIG improvement over Population Finder.

I am "100% European" now: 53% European Coastal Plain, 43% European Coastal Isles, and 4% North Circumpolar.

The myOrigins display is much more appealing than the old PF, too.

I am interested in finding out more about the populations that were used for comparison.

[engisdene]

Could be Turkish or from the Caucasus, depends on what that green spot in Turkey represents in the breakdown. The Mediterranean/Northeast Europe/East Asian looks about right for someone from the Turkey-Caucasus region, maybe even Ukraine.

Yes, it is a Turkish individual result.

[bbmoon]

So does this mean ftdna will still use 4% and above confidence for giving one a population match in their criteria?

[engisdene]

So does this mean ftdna will still use 4% and above confidence for giving one a population match in their criteria?

1% exist someones results.

[Germanica]

I had a look at the new "myOrigins" thingy this morning, and I like it. IMHO, it is a BIG improvement over Population Finder.

I am "100% European" now: 53% European Coastal Plain, 43% European Coastal Isles, and 4% North Circumpolar.

The myOrigins display is much more appealing than the old PF, too.

I am interested in finding out more about the populations that were used for comparison.

Where are you seeing this? I see nothing new or different in my FTDNA dashboard/homepage. Nothing that says myOrigins, it still says Population Finder and when I click on it, it's still the same display. Sounds like they are maybe rolling it out slowly in batches, so not everyone has the new display yet.

[LynCra]

It seems that only project admins have access at the moment.

[MoberlyDrake]

It says "Secret beta". Are we allowed to make any comments at all? To what extent and what kind of things should we report?

[rod]

Indeed! As author of the original Pop Finder, I'd sure love
a peek. I promise not to blab anything to anybody, even FTDNA.

Doug McDonald
[email protected]

After seeing the "new improved", I've developed a greater appreciation for the old Population Finder.

[MMaddi]

It seems that only project admins have access at the moment.

That's right. The new myOrigins data is available to project administrators for members of their project, through the Group Administrator Page. As Sicily Project administrator, I can view the new myOrigins results for the approximately 190 members who've tested Family Finder.

Since this is in beta, it seems FTDNA is wisely seeking the input of project administrators about any problems with the results or opinions about the accuracy of or any possible weaknesses in the percentages given. I plan to look more closely over the next several days at the results and offer my opinion of their accuracy or inaccuracy.

I'll also post here if I find any interesting trends, at least as far as it relates to Sicilian ancestry. I'm particularly interested in what myOrigins gives to project members who were given 100% Middle Eastern (Jewish) in Population Finder. That was a major problem in Population Finder for Sicily Project members. Hopefully, the myOrigins algorithm will give those members more plausible admixture percentages.

[MoberlyDrake]

As a member of the Sicilian project who was always aggravated with getting a questionnaire asking which Middle Eastern population my family members belonged to every time I tried to look at Population Finder, I still see a considerable amount of Middle Eastern for my family members.

[MoberlyDrake]

After seeing the "new improved", I've developed a greater appreciation for the old Population Finder.

I think you may be right!!!

[MorrigensCrow]

It is nice.

I am so jealous right now! So, if you don't mind me asking Engisdene, is the whopping 74% ME accurate for you? I have read a lot of complaints from people about their small amounts of unexpected Middle Eastern DNA in the PF, but yours is quite high, so I assume you have actual ME ancestry?

Personally, I have 12+% ME and 87+% Orkney (Old PF), and while I have yet to find any documentation of ME ancestry, I actually look like I could be part Middle Eastern or maybe Mediterranean, and that is very odd. My mom's DNA tests (through another company) pointed her to India, so we are baffled about that too. Supposedly, we have some distant Native American ancestry, but there is no proof, yet my mother looks very Native American, so does has anyone ever heard of Native Americans having ME DNA markers?

[Osiris]

I do like it a lot. The only part that I was disappointed in was my mother's Polynesian DNA. It still reads as East Asia. Only Ancestry.com calls it Polynesian, 23andme also calls it East Asian. The story talks about the East Asians migrating out into Madagascar, Oceania all the way out to Easter Island but the Heat Map is centered on South China and Madagascar. My ancestors haven't lived there in several thousand years so there's still lots of room for improvement in this group. I assume it's just their samples are wanting in this group but I was hoping the map would stretch out into the Pacific at least.

The Middle East % did drop a lot.

[mkdexter]

I like the heat map approach however, I'm not sure I would go for modern country names as a background map. It would be neat to see a version with the matching region names instead.

The famous "Orcadian" is now gone replaced with European Costal Islands, with some of it being separated out to include what they call European Northlands and European Costal Plain as well. So if you had "Orcadian" before, you will probably have "European Costal Islands" now with some added amounts from the neighboring locations.

Overall I would call it definitely an improvement. It certainly lines up with what I know from other tests and reporting. My 8% North African (PFv1) is now more accurately depicted as 2% African like the other company and McDonald reports show.



Matt.

[McNinja]

Here's a look at my new results:

First the old PF: http://i.imgur.com/qfte0Qf.png

New PF:

Central/South Asian
- Indian Tectonic 38%
- Eurasian Heartland 49%

European
- North Circumpolar 6%
- Trans-Ural Peneplain 3%

East Asian
- East Asian Coastal Islands 5%

The old one is still more accurate or relevant for me (including Dr. McDonald's own analysis) but the new one isn't terrible. Better than 23andMe's Ancestry Composition and Geno 2.0 as well.

I'm curious about Indian Tectonic though. Most South Asians on these genome sites like myself are interested in the history of Ancestral North and South Indian, what separates the West Eurasian from East Eurasian, etc but FTDNA's Indian components don't even try to tackle such a task (perhaps wisely). They include Indo-Aryan (North Indian) in the Indian Tectonic component. So Indian Tectonic is a snapshot of India like 2000 years ago, not more than 4000 years ago.

Eurasian Heartland is also very interesting and I'd love to know more about it, like what populations they used. The Orcadian that some of us South Asians were getting has turned into Europeans now getting some Eurasian Heartland instead which is more anthropologically/historically accurate. Most theorized this link between the Gedrosia/South Asia region and Northwestern Europe (Scotland, Orkney Islands) was a result of some specific ancient migration and their description seems to imply as much with the mention of Sarmatians and Scythians.

This is also the first time a test has managed to pull more of my ancestry outside of India. That's a good thing because everyone else (including my mother) of my ethnic group registers more Gedrosian/South Central Asian than Indian usually except me. And my father is a second cousin of my mother and their family's pretty inbred so me getting all sorts of Siberian/Beringian/Papuan/Native American noise and apparently a lot more European IBS/IBD matches than my mom, yet higher Indian in admixture was weird. So I'm really curious what they did.

I do miss the margin of error estimation. It kind of showed that these numbers weren't set in stone and were estimates.