PDA

View Full Version : S/SC ASIAN Results :: MyHeritageDNA



Censored
08-09-2018, 12:44 AM
Moderator Note: Post your MyHeritageDNA results here. At your convenience add your results to this spreadsheet: S/SC ASIAN :: myHertitageDNA Ancestry Composition {link not available yet}.
DO NOT create a new thread for individual results

Censored Original Post

Post yours.

Here's mine:
25072

Edit: It was a 23andme transfer by the way.

MonkeyDLuffy
08-09-2018, 01:41 AM
Mine was imported from 23&me:

25073

misanthropy
08-09-2018, 05:07 AM
23andme v4 transfer. I don't even remember uploading here, so I guess it was from a while back.

https://i.imgur.com/fv0uHG3.png

Censored
08-09-2018, 05:22 AM
23andme v4 transfer. I don't even remember uploading here, so I guess it was from a while back.

https://i.imgur.com/fv0uHG3.png

I'm surprised you got Africa and East Asia but not west Asia. Strange.

midichlorian
08-09-2018, 06:19 AM
Gotta wait for them to accept V5 chip :(

MuslimPatel123
08-09-2018, 08:34 AM
I get 4.2% Ashkenazi Jewish. Jewish estimates are probably inflated since this is an Israeli company.

https://i.imgur.com/Lp2UAPd.png

Lollybolly
08-09-2018, 12:30 PM
Here are mine. I get a significant amount of Caucasian DNA in this one (near 20%)
25099

kush
08-09-2018, 12:35 PM
100% Desi

https://i.imgur.com/bfOAZLf.png

Censored
08-09-2018, 03:58 PM
100% Desi

https://i.imgur.com/bfOAZLf.png

Unsurprising, I had a feeling they used some South Indian groups as reference populations for the South Asian category since I got such a high percentage. Though they probably still didn’t capture the full diversity of South India with me getting .8% Papuan.

Censored
08-09-2018, 04:05 PM
Here are mine. I get a significant amount of Caucasian DNA in this one (near 20%)
25099

You’re always getting Middle Eastern. I always feel upset when I see my results mainly because I was misled. Feelsbadman.

Lollybolly
08-09-2018, 04:15 PM
You’re always getting Middle Eastern. I always feel upset when I see my results mainly because I was misled. Feelsbadman.

You should be proud to be "pure" Indian dude!

Censored
08-09-2018, 07:55 PM
You should be proud to be "pure" Indian dude!

Yeah it wouldn't be as bad if I had just known from the start, but I had many reasons to believe otherwise. Sometimes I want to tell my dad the truth about his ancestry.

misanthropy
08-09-2018, 08:18 PM
I'd still take this calc with a decent sized grain of salt. It doesn't pick up ancestries that repeatedly show up on other calcs for me at least.

bmoney
08-10-2018, 12:17 AM
Yeah it wouldn't be as bad if I had just known from the start, but I had many reasons to believe otherwise. Sometimes I want to tell my dad the truth about his ancestry.

@Censored, dont take it personally man its not just you

Its very socio-political in nature, all South Asians have some level of identity issues, especially South Asian Muslims for various historical and political reasons

Soon Hindi-belt North Indians will be feeling the identity pain too after Rakhigarhi

Dravidians like me have never known what we are, have no written history lol

Censored
08-10-2018, 12:54 AM
@Censored, dont take it personally man its not just you

Its very socio-political in nature, all South Asians have some level of identity issues, especially South Asian Muslims for various historical and political reasons

Soon Hindi-belt North Indians will be feeling the identity pain too after Rakhigarhi

Dravidians like me have never known what we are, have no written history lol

Yeah, I'm well aware of it. I don't want to turn into that "stereotype" of the Indian who thinks they are/wants to be something they're not, but I really felt I had a much stronger justification for believing there was some kind of possible mix in my ancestry than the vast majority of Indians who believe the same thing. I've seen people claiming really bizarre and quite frankly impossible things like Hindus saying they are part Persian or shit like that and thought that if anything, I would be the one having it given my dad's family being Hyderabadi Muslims, and many of his extended family displaying obvious Afghan or west Asian admixture.

But families are really messy, and I guess my dad's own line is just South Indian, even if more distant relatives are mixed with something exotic.

And hey, at least you Hindus know your caste lol. I still have to go through the painstaking process of figuring out what my parents must have been before converting.

MonkeyDLuffy
08-10-2018, 01:36 AM
Yeah, I'm well aware of it. I don't want to turn into that "stereotype" of the Indian who thinks they are/wants to be something they're not, but I really felt I had a much stronger justification for believing there was some kind of possible mix in my ancestry than the vast majority of Indians who believe the same thing. I've seen people claiming really bizarre and quite frankly impossible things like Hindus saying they are part Persian or shit like that and thought that if anything, I would be the one having it given my dad's family being Hyderabadi Muslims, and many of his extended family displaying obvious Afghan or west Asian admixture.

But families are really messy, and I guess my dad's own line is just South Indian, even if more distant relatives are mixed with something exotic.

And hey, at least you Hindus know your caste lol. I still have to go through the painstaking process of figuring out what my parents must have been before converting.

Technically we're part Iranians, ancient Iranians tbh. Iran N is big chunk of our ancestry afterall lol. Yea it's hard for south asian Muslims to find out their real heritage, especially if they converted long time ago. Syed claiming communities get the hardest hits since all of them have family tree connecting to founder of Islam I believe, with rest of records forgotten.

You do show very minor Levant farmer like mix in nmonte, so hey, there is your link. It's very small but it still backs up the family claim of somewhat exotic ancestry.

I've asked a lot of questions about my own paternal side over the years from my relatives. Since I was not told much about it, because of all the moving my grandfather and his family did. I still score like an outlier for a Punjabi Tarkhan based on the samples we've, so hey we all got some mystery mix somewhere.

khanabadoshi
08-10-2018, 02:09 AM
https://i.gyazo.com/9f76dbf51e25c4b36adbd74b78b1a6ac.png

khanabadoshi
08-10-2018, 02:22 AM
I'll make this the official thread for myHeritageDNA and make that spreadsheet public. At some point LOL.

Censored
08-10-2018, 02:36 AM
Technically we're part Iranians, ancient Iranians tbh. Iran N is big chunk of our ancestry afterall lol. Yea it's hard for south asian Muslims to find out their real heritage, especially if they converted long time ago. Syed claiming communities get the hardest hits since all of them have family tree connecting to founder of Islam I believe, with rest of records forgotten.

You do show very minor Levant farmer like mix in nmonte, so hey, there is your link. It's very small but it still backs up the family claim of somewhat exotic ancestry.

I've asked a lot of questions about my own paternal side over the years from my relatives. Since I was not told much about it, because of all the moving my grandfather and his family did. I still score like an outlier for a Punjabi Tarkhan based on the samples we've, so hey we all got some mystery mix somewhere.

I'd still say Persian is a far cry from any kind of ancient Iranian whether speaking of Iranian N farmers or Indo-Iranians.

Someone did an nMonte run using Saidu Sharif, Paniya, and Velamas and got a fit of 1.6 for me. On top of that I recently tried running nMonte with Iyers and Paniya to see which modern pops could approximate my ancestry closest, and got a fit of 1.45. In both scenarios adding a West Asian pop. only made the fit worse.

I remember at one point you said that it seemed that one side of my family seemed upper caste and the other tribal-like; I am increasingly partial to that theory after everything I have seen so far.

MonkeyDLuffy
08-10-2018, 02:50 AM
I'd still say Persian is a far cry from any kind of ancient Iranian whether speaking of Iranian N farmers or Indo-Iranians.

Someone did an nMonte run using Saidu Sharif, Paniya, and Velamas and got a fit of 1.6 for me. On top of that I recently tried running nMonte with Iyers and Paniya to see which modern pops could approximate my ancestry closest, and got a fit of 1.45. In both scenarios adding a West Asian pop. only made the fit worse.

I remember at one point you said that it seemed that one side of my family seemed upper caste and the other tribal-like; I am increasingly partial to that theory after everything I have seen so far.

It's your steppe in comparison to other neighbouring communities that makes me wonder if you've some Brahmin like mix or ancestry from one side. Checking your family members or even Grandparents will solve it.

Censored
08-10-2018, 02:53 AM
It's your steppe in comparison to other neighbouring communities that makes me wonder if you've some Brahmin like mix or ancestry from one side. Checking your family members or even Grandparents will solve it.

Hmm, how much steppe do I appear to have versus neighboring communities?

If it's confirmed that I appear partly Brahmin like then I can already pretty much tell you that it's from my father.

Sapporo
08-11-2018, 05:48 AM
@khana

Something must be off. Some of your family is scoring low West Asian and high Euro and vice versa for me. I'm presuming your results are 23andMe or FTDNA transfers as well?

25175

edit:

I added my Ancestry DNA transfer as well. Numbers changed slightly. 0.5% to a 1% here and there.


25200

khanabadoshi
08-11-2018, 06:50 AM
@khana

Something must be off. Some of your family is scoring low West Asian and high Euro and vice versa for me. I'm presuming your results are 23andMe or FTDNA transfers as well?

25175

Yeah, they are 23andme V4 transfers and one FTDNA. Maybe they changed? I haven't looked at myHertiageDNA in a while. Let me check.

Sapporo
08-11-2018, 08:38 AM
Yeah, they are 23andme V4 transfers and one FTDNA. Maybe they changed? I haven't looked at myHertiageDNA in a while. Let me check.

MyHeritage let me upload my 23andMe V3 and Ancestry DNA at the same time. I'm not even sure which one they used to be honest. But you don't find the admixture breakdown weird? You and some of your family (brother and sister as well) should be scoring higher West Asian than me and lower Euro based on consistency with other calcs or runs but on here it's the opposite. There is something off when I'm scoring the same West Asian % as Lollybolly when I usually score much less Caucasus or West Asian like admixture (outside of the Iran N/Baloch like components).

Another example is your Multani Syed/Durrani friend. 0% West Asian and Central Asian but just over 15% North Euro (although he gets 6% Middle Eastern)? Your brother is also scoring more South Asian than me and only 1.7% West Asian but a whopping 16.8% North Euro and 4.4% Ashkenazi Jew?

Lollybolly
08-11-2018, 10:25 AM
This calculator is shitty, just like DNA.Land. Better stick to GEDmatch

Jatt1
08-19-2018, 08:27 PM
Jatt1 Myheritage Results

Strange results?

Asia
90.6%
South Asia
65.8%
South Asian
65.8%
West Asia
20.6%
West Asian
20.6%
Central Asia
4.2%
Central Asian
4.2%
Europe
9.4%
North and West Europe
6.8%
Finnish
4.6%
Irish, Scottish, and Welsh
1.3%
Scandinavian
0.9%
East Europe
2.6%
Baltic
2.6%

HarappaWorld Oracle results:

Admix Results (sorted):

# Population Percent
1 Baloch 38.78
2 S-Indian 30.45
3 Caucasian 13.02
4 NE-Euro 11.8
5 SW-Asian 2.12
6 Siberian 1.54
7 Beringian 1.12
8 Mediterranean 0.86
9 American 0.25
10 Papuan 0.06
11 San 0.01

Single Population Sharing:

# Population (source) Distance
1 punjabi-khatri (harappa) 2.94
2 kashmiri-pandit (reich) 3.87
3 punjabi-jatt-sikh (harappa) 4.12
4 punjabi-jatt-muslim (harappa) 4.38
5 punjabi (harappa) 4.63
6 punjabi-brahmin (harappa) 4.67
7 kashmiri (harappa) 4.76
8 sindhi (harappa) 5.09
9 singapore-indian-c (sgvp) 5.48
10 kashmiri-pahari (harappa) 6.54
11 punjabi-ramgarhia (harappa) 6.86
12 punjabi-arain (xing) 7.34
13 pathan (hgdp) 7.54
14 up-muslim (harappa) 7.61
15 gujarati-muslim (harappa) 7.75
16 haryana-jatt (harappa) 8.08
17 bhatia (harappa) 8.61
18 nepalese-a (xing) 8.77
19 sindhi (hgdp) 9.34
20 burusho (hgdp) 10.35

Mixed Mode Population Sharing:

# Primary Population (source) Secondary Population (source) Distance
1 75.5% punjabi-brahmin (harappa) + 24.5% pashtun (harappa) @ 2.01
2 83.1% punjabi-khatri (harappa) + 16.9% up-brahmin (harappa) @ 2.02
3 75% pathan (hgdp) + 25% srivastava (reich) @ 2.04
4 90.6% punjabi-brahmin (harappa) + 9.4% stalskoe (xing) @ 2.05
5 64.6% pathan (hgdp) + 35.4% brahmin-uttar-pradesh (metspalu) @ 2.07
6 59.1% pathan (hgdp) + 40.9% up-brahmin (harappa) @ 2.07
7 80.3% punjabi-khatri (harappa) + 19.7% nepalese-a (xing) @ 2.07
8 86.5% punjabi-khatri (harappa) + 13.5% brahmin-uttar-pradesh (metspalu) @ 2.09
9 68.1% pathan (hgdp) + 31.9% vaish (reich) @ 2.11
10 90.8% punjabi-brahmin (harappa) + 9.2% urkarah (xing) @ 2.12
11 76.3% punjabi-khatri (harappa) + 23.7% punjabi-ramgarhia (harappa) @ 2.13
12 67.3% punjabi-khatri (harappa) + 32.7% punjabi-brahmin (harappa) @ 2.13
13 89.1% singapore-indian-c (sgvp) + 10.9% urkarah (xing) @ 2.15
14 91.4% punjabi-brahmin (harappa) + 8.6% kumyk (yunusbayev) @ 2.15
15 71.6% punjabi-khatri (harappa) + 28.4% singapore-indian-c (sgvp) @ 2.15
16 88.6% punjabi-khatri (harappa) + 11.4% vaish (reich) @ 2.16
17 91.1% punjabi-brahmin (harappa) + 8.9% lezgin (behar) @ 2.19
18 89.2% punjabi-khatri (harappa) + 10.8% up-kshatriya (metspalu) @ 2.21
19 84.3% pathan (hgdp) + 15.7% bhil (reich) @ 2.22
20 90.3% singapore-indian-c (sgvp) + 9.7% chechen (yunusbayev) @ 2.22


MDLP K16 Modern Oracle results:

Admix Results (sorted):

# Population Percent
1 Indian 49.08
2 Caucasian 25.49
3 Steppe 11.56
4 SouthEastAsian 3.48
5 NorthEastEuropean 3.23
6 Arctic 2.21
7 Oceanic 1.58
8 Siberian 1.2
9 NearEast 1.07
10 Neolithic 0.58
11 Australian 0.21
12 Ancestor 0.17
13 Amerindian 0.12

Single Population Sharing:

# Population (source) Distance
1 Pathan (Punjab) 3
2 Kashmiri_Pandit (Kashmir) 4.28
3 Jatt (Pahari) 4.38
4 Gujjar (Punjab) 5.63
5 Jatt (Muslim) 6.65
6 Jatt (Haryana) 7.28
7 Pashtun (Pakistan) 8.23
8 Burusho (Pakistan) 8.25
9 Brahmin (India) 9.1
10 Sindhi (Sindh) 10.13
11 Meena (Rajasthan) 10.18
12 Balochi (Baluchistan) 10.56
13 Brahui (Baluchistan) 10.82
14 Jew (Mumbai) 10.88
15 Kshatriya (India) 11.41
16 Brahmin (Uttar_Pradesh) 12.46
17 Vaish (Odisha) 12.75
18 GujaratiA (Gujarat) 13.61
19 Jew (Cochin) 13.89
20 Makrani (Pakistan) 14

Mixed Mode Population Sharing:

# Primary Population (source) Secondary Population (source) Distance
1 80.7% Gujjar (Punjab) + 19.3% Shugnan (Badachshan) @ 2.38
2 81.1% Gujjar (Punjab) + 18.9% Rushanvanch (Gorno-Badakhshan) @ 2.39
3 78.8% Gujjar (Punjab) + 21.2% Ishkasim (Gorno-Badakhshan) @ 2.47
4 80.4% Gujjar (Punjab) + 19.6% Tajik (Pomiri_Tajikistan) @ 2.47
5 98.6% Pathan (Punjab) + 1.4% Itelmen (Kamchatka) @ 2.6
6 76.2% Gujjar (Punjab) + 23.8% Pashtun (Afghanistan) @ 2.61
7 98.5% Pathan (Punjab) + 1.5% Koryak (Koryak_Okrug) @ 2.62
8 71.1% Pathan (Punjab) + 28.9% Kashmiri_Pandit (Kashmir) @ 2.67
9 98.6% Pathan (Punjab) + 1.4% Murut (Borneo) @ 2.71
10 98.6% Pathan (Punjab) + 1.4% Luzon (Philippines) @ 2.72
11 98.7% Pathan (Punjab) + 1.3% Ami (Taiwan) @ 2.72
12 98.6% Pathan (Punjab) + 1.4% Dusun (Borneo) @ 2.72
13 98.7% Pathan (Punjab) + 1.3% Atajal (Taiwan) @ 2.72
14 98.6% Pathan (Punjab) + 1.4% Lebbo (Borneo) @ 2.73
15 98.7% Pathan (Punjab) + 1.3% Igorot (Philippines) @ 2.73
16 98.5% Pathan (Punjab) + 1.5% Cambodian (Cambodia) @ 2.73
17 98.7% Pathan (Punjab) + 1.3% Dai (Xishuangbanna) @ 2.74
18 98.6% Pathan (Punjab) + 1.4% Vizayan (Borneo) @ 2.74
19 98.4% Pathan (Punjab) + 1.6% Thai (Thailand) @ 2.74
20 98.6% Pathan (Punjab) + 1.4% Filipino (Philippines) @ 2.74

aaronbee2010
08-20-2018, 12:48 AM
MyHeritage DNA sample, no transfers.

https://i.gyazo.com/a00091ef2becc56c5294b7b0cea4dc6a.png

Dr_McNinja
10-17-2018, 04:22 AM
https://docs.google.com/spreadsheets/d/1XGY-UIBC2GcUWGl1FizNvu9ofle4ZPHvfNmbrVYxGMo/edit#gid=899259100

Dr_McNinja
10-17-2018, 05:11 AM
The best part about this site is the animated movie of your results with the music

Mingle
11-16-2018, 05:34 PM
23andMe V5 transfer

https://i.imgur.com/RQA2vV8.png?1

Adam A
12-26-2018, 04:55 PM
27945

Amber29
12-26-2018, 05:59 PM
Original results which seems to be close to AncestryDNA Transfer too.
27953

My Fathers (ancestryTransfer):
South asian: 70.2
West asian: 22.0
Europe: 7.8

wasnt there a thread for this already or is that removed? lol

MonkeyDLuffy
12-26-2018, 06:19 PM
Original results which seems to be close to AncestryDNA Transfer too.


My Fathers (ancestryTransfer):
South asian: 70.2
West asian: 22.0
Europe: 7.8

wasnt there a thread for this already or is that removed? lol

You might want to take off your real name from there.

khanabadoshi
12-27-2018, 05:26 AM
Original results which seems to be close to AncestryDNA Transfer too.
27953

My Fathers (ancestryTransfer):
South asian: 70.2
West asian: 22.0
Europe: 7.8

wasnt there a thread for this already or is that removed? lol

This is the main thread for MyHeritage. There is a spreadsheet for it on the first page, you can add your relative's results.
EDIT: Looks like I didn't make a public spreadsheet yet. I should do that....

Dr_McNinja
01-23-2019, 01:49 AM
https://docs.google.com/spreadsheets/d/1XGY-UIBC2GcUWGl1FizNvu9ofle4ZPHvfNmbrVYxGMo/edit#gid=899259100

Mostly all 23andMe/FTDNA imports

Amber29
01-23-2019, 05:24 PM
https://docs.google.com/spreadsheets/d/1XGY-UIBC2GcUWGl1FizNvu9ofle4ZPHvfNmbrVYxGMo/edit#gid=899259100

Mostly all 23andMe/FTDNA imports

Woaaaaah that alot pages and list lol i kept pressing next next next - did everyone add them selves in or gotta ask to be put in?

Jatt1
01-23-2019, 09:20 PM
https://docs.google.com/spreadsheets/d/1XGY-UIBC2GcUWGl1FizNvu9ofle4ZPHvfNmbrVYxGMo/edit#gid=899259100

Mostly all 23andMe/FTDNA imports

What calculator you uses for that spreadsheets, it certainly isn't Harappa?

aaronbee2010
01-23-2019, 10:19 PM
What calculator you uses for that spreadsheets, it certainly isn't Harappa?

I think it's just the MyHeritage results straight from the website.

Taqi_G
01-31-2019, 12:23 AM
Myheritage:

South asia 83,9%

West asia 3,7%

Scandinavian 6,2%

Finnish 2,7%

Ashkenazi Jew 2,3%

Central America 1,2%

scobar
03-04-2019, 01:27 AM
MyHeritage results (23andMe v5 upload)

29146

scobar
05-10-2019, 07:41 AM
MyHeritage results (23andMe v5 upload)

29146

Ancestry V2 upload

Pretty much identical to v5

30359

prashantvaidwan
09-02-2019, 10:16 AM
One of my known is tested on my heritage, got his results but could not locate his Raw dna file. Any idea?

Amber29
09-02-2019, 11:44 AM
One of my known is tested on my heritage, got his results but could not locate his Raw dna file. Any idea?

Manage DNA kits and on the rigyt hand side it has a three dot thingy click on that and dowload raw data.

Dr_McNinja
12-22-2019, 04:13 PM
May not be the best calculator, but everyone I know getting tested at 23andMe is paying to port their results to MH just for the animation breakdown, lol. They win marketing.

MonkeyDLuffy
01-29-2020, 12:17 PM
Noticed they're charging for ethnicity estimate now if you did DNA transfer. $39 CAD. Hmm, I can get nmonte for my mum and dad in that lol.

Censored
01-29-2020, 12:21 PM
May not be the best calculator, but everyone I know getting tested at 23andMe is paying to port their results to MH just for the animation breakdown, lol. They win marketing.

Why, so they can tell us that we are 20% west Asian and 18% Euro even though we know we’re not. Lol

Sapporo
01-31-2020, 03:26 AM
Why, so they can tell us that we are 20% west Asian and 18% Euro even though we know we’re not. Lol

I'm not an expert on MyHeritage's algorithim but I believe they look at more ancient DNA than 23andMe in comparison (which only looks at the last 500 years or so).

Censored
01-31-2020, 04:24 AM
I'm not an expert on MyHeritage's algorithim but I believe they look at more ancient DNA than 23andMe in comparison (which only looks a the last 500 years or so).

which is a problem with these steppe MLBA groups are 70-80% similar to anglos/nordics

Sapporo
01-31-2020, 06:29 AM
which is a problem with these steppe MLBA groups are 70-80% similar to anglos/nordics

Even closer to some Baltic/Finnic/Uralic groups. The point being that the algorithim isn't distinguishing between moderns and ancients. So, it should denote that. However, the "Euro" and "West Asian" being picked up probably has some relation to Steppe or BMAC/Iran-N ancestry for South Asians so it's not entirely wrong. It's just the presentation of the data. Even the South Asian component on MyHeritage is probably based on some modern South Asian pop or user data submitted to the company.

Censored
01-31-2020, 06:43 AM
Even closer to some Baltic/Finnic/Uralic groups. The point being that the algorithim isn't distinguishing between moderns and ancients. So, it should denote that. However, the "Euro" and "West Asian" being picked up probably has some relation to Steppe or BMAC/Iran-N ancestry for South Asians so it's not entirely wrong. It's just the presentation of the data. Even the South Asian component on MyHeritage is probably based on some modern South Asian pop or user data submitted to the company.

That's where it gets tricky, when you go far that back in time it gets hard. It's not just that steppe MLBA is similar to those groups, it is essentially one of them as it shares common ancestry via CWC.

Dr_McNinja
01-31-2020, 09:30 PM
MyHeritage is similar to some of the Gedmatch calculators, they just don't have a Gedrosian/Baloch component so admixture gets split between South Asia and West Asia which is less stable than having a Gedrosian component.

They do have a Nepali component but I don't know anyone who's gotten that yet. Have any of the users from Nepal here gotten tested on MyHeritage? The music is cool.

FTDNA has two components, but their "Central Asia" which seems like North India/Pakistan seems a little strange. But they said they're changing the myOrigins soon in a recent email.

Btw, if you haven't seen/heard all of them yet:

https://drive.google.com/open?id=19qjWfuyZGO8LLeGhyKU6vU53UJoxfecT

https://drive.google.com/open?id=1cf5r7bitS5_EpyWWMz2pSTDb0MA7D8LL

https://drive.google.com/open?id=1mkiVYJ16gGLk9oAMn_hvQZQ98iZsjpHV

There's also a lot of people posting their MH results on YouTube because of the animation, including South and Central Asians.

Dr_McNinja
01-31-2020, 10:01 PM
Another quirk is that some South Asians' European score gets inflated with West Asian admixture and vice-versa. My brother in law, HRP0402, had like 15-16% South Euro because all his West Asian disappeared whereas everyone else like him had most of that in West Asian. Same for my family aside from me. They were averaging 12-13% Euro with very little West Asian compared to other Punjabi Jatts.

South Asian component also has some interference with Euro which is a little odd. There haven't been a lot of these outliers but I've seen 1 or 2 of each at least out of a few dozen results.

Amber29
01-31-2020, 10:12 PM
MyHeritage is similar to some of the Gedmatch calculators, they just don't have a Gedrosian/Baloch component so admixture gets split between South Asia and West Asia which is less stable than having a Gedrosian component.

They do have a Nepali component but I don't know anyone who's gotten that yet. Have any of the users from Nepal here gotten tested on MyHeritage? The music is cool.

FTDNA has two components, but their "Central Asia" which seems like North India/Pakistan seems a little strange. But they said they're changing the myOrigins soon in a recent email.

Btw, if you haven't seen/heard all of them yet:

https://drive.google.com/open?id=19qjWfuyZGO8LLeGhyKU6vU53UJoxfecT

https://drive.google.com/open?id=1cf5r7bitS5_EpyWWMz2pSTDb0MA7D8LL

https://drive.google.com/open?id=1mkiVYJ16gGLk9oAMn_hvQZQ98iZsjpHV

There's also a lot of people posting their MH results on YouTube because of the animation, including South and Central Asians.

I posted mine on Youtube xD hahaha when i first did mine

Dr_McNinja
01-31-2020, 10:23 PM
https://docs.google.com/spreadsheets/d/1XGY-UIBC2GcUWGl1FizNvu9ofle4ZPHvfNmbrVYxGMo/edit#gid=899259100

Dr_McNinja
02-05-2020, 01:41 AM
Updated my list with a lot of people:

https://docs.google.com/spreadsheets/d/1XGY-UIBC2GcUWGl1FizNvu9ofle4ZPHvfNmbrVYxGMo/edit#gid=899259100

The two samples at the bottom under the names Bashyal and 'Arul Paul' are super interesting. Is Bashyal from Nepal?

pnb123
02-05-2020, 02:12 AM
Updated my list with a lot of people:

https://docs.google.com/spreadsheets/d/1XGY-UIBC2GcUWGl1FizNvu9ofle4ZPHvfNmbrVYxGMo/edit#gid=899259100

The two samples at the bottom under the names Bashyal and 'Arul Paul' are super interesting. Is Bashyal from Nepal?
Yeah Bashyal is most likely from Nepal.

Sapporo
02-05-2020, 04:39 AM
Yeah Bashyal is most likely from Nepal.

How would you explain 24.7% West Asian and 3.5% Euro combo for a Nepali? That doesn't look like a Bahun to me.

@Dr_McNinja

I'm positive the Batra is a Hindu or Sikh Arora. Also, I seem to score inflated West Asian on MyHeritage (at least for my group) as I'm not too far off Lollybolly or your Arain friend as well as other Punjabi Khatri/Arora or Lohana types. What do you think is causing that?

agent_lime
02-05-2020, 04:54 AM
How would you explain 24.7% West Asian and 3.5% Euro combo for a Nepali? That doesn't look like a Bahun to me.

@Dr_McNinja

I'm positive the Batra is a Hindu or Sikh Arora. Also, I seem to score inflated West Asian on MyHeritage (at least for my group) as I'm not too far off Lollybolly or your Arain friend as well as other Punjabi Khatri/Arora or Lohana types. What do you think is causing that?

Can confirm Batra is an Arora from Multan(probably) last name.

MonkeyDLuffy
02-05-2020, 05:26 AM
How would you explain 24.7% West Asian and 3.5% Euro combo for a Nepali? That doesn't look like a Bahun to me.

@Dr_McNinja

I'm positive the Batra is a Hindu or Sikh Arora. Also, I seem to score inflated West Asian on MyHeritage (at least for my group) as I'm not too far off Lollybolly or your Arain friend as well as other Punjabi Khatri/Arora or Lohana types. What do you think is causing that?

There's something wrong with their calculator. Ramgarhia PG scores almost 20% european on it. And another Tarkhan sample has 22% west asian.

Dr_McNinja
02-05-2020, 06:22 AM
They're using moderns so it seems West Asia has some European in it. As it goes up, European goes down and vice-versa. But it's still West Asia so some people get a really strong pull to it (probably with high Baloch numbers in HarappaWorld).

Central Asia likely has some European too.

Also depends on the number of markers they use and the platform you were tested on.

Also, South Asia must have European/Central Asia and all that within it too because it's so large.

Basically it's like an unstable HarappaWorld where you have an S-Indian/Baloch component (and S-Indian has WSHG) and a Baloch/Caucasian/WSHG component (West Asia + Central Asia). And the European components have Caucasian and probably some Baloch too of course.

When components overlap like that, whether in Admixture or an nMonte/Oracle-type analysis, minor fluctuations can become amplified. So if you have a stronger draw to something in West Asia or Central Asia (so even the slightest pull to Baloch/Caucasian in Harappa for example), they'll pull in a bunch of your other admixture, but if not then it will go into European because it's like left out. South Asia can also suck in West Asian/European it seems but to a lesser degree.

But there's something else going on which is probably whatever their own method was to set this up in terms of markers/populations/datasets they used. That last bit is unique to them. I have no idea how their test is affected by taking their DNA test vs. autosomal transfer. LivingDNA showed a big difference. I think it might be less as important here because they used FTDNA's lab for so long but who knows.

So like my family were 80-90% South Asian and 8-14% European, and skipping over the West Asian-heavy Jatts (where Sapporo landed), you find another batch of Jatts who are basically less South Asian versions of us (70-80% South Asian, 11-19% Euro). The Jatts who had probably more Baloch affinity got sucked into the West Asian-heavy area.

pnb123
02-05-2020, 06:28 AM
How would you explain 24.7% West Asian and 3.5% Euro combo for a Nepali? That doesn't look like a Bahun to me.

@Dr_McNinja

I'm positive the Batra is a Hindu or Sikh Arora. Also, I seem to score inflated West Asian on MyHeritage (at least for my group) as I'm not too far off Lollybolly or your Arain friend as well as other Punjabi Khatri/Arora or Lohana types. What do you think is causing that?
People who use that surname belong to either Bahun, Chhetri, or Dalit (Sarki) caste. But this is just my internet knowledge. I actually haven't met a single person who uses that surname ever in my life. Maybe poi bro knows more about it. Could that person belong to Dalit caste (going by low European %)?

Dr_McNinja
02-05-2020, 06:45 AM
It seems to handle Afghans, high-Steppe Jats (Haryana/UP), and high-AASI/S-Indian individuals well (with comparable results to other calculators). It has trouble with the area between South Asia/West Asia. But the results still more or less average out (if you were to graph these) and follow their own strange but consistent pattern.

Censored
02-05-2020, 09:09 AM
They're using moderns so it seems West Asia has some European in it. As it goes up, European goes down and vice-versa. But it's still West Asia so some people get a really strong pull to it (probably with high Baloch numbers in HarappaWorld).

Central Asia likely has some European too.

Also depends on the number of markers they use and the platform you were tested on.

Also, South Asia must have European/Central Asia and all that within it too because it's so large.

Basically it's like an unstable HarappaWorld where you have an S-Indian/Baloch component (and S-Indian has WSHG) and a Baloch/Caucasian/WSHG component (West Asia + Central Asia). And the European components have Caucasian and probably some Baloch too of course.

When components overlap like that, whether in Admixture or an nMonte/Oracle-type analysis, minor fluctuations can become amplified. So if you have a stronger draw to something in West Asia or Central Asia (so even the slightest pull to Baloch/Caucasian in Harappa for example), they'll pull in a bunch of your other admixture, but if not then it will go into European because it's like left out. South Asia can also suck in West Asian/European it seems but to a lesser degree.

But there's something else going on which is probably whatever their own method was to set this up in terms of markers/populations/datasets they used. That last bit is unique to them. I have no idea how their test is affected by taking their DNA test vs. autosomal transfer. LivingDNA showed a big difference. I think it might be less as important here because they used FTDNA's lab for so long but who knows.

So like my family were 80-90% South Asian and 8-14% European, and skipping over the West Asian-heavy Jatts (where Sapporo landed), you find another batch of Jatts who are basically less South Asian versions of us (70-80% South Asian, 11-19% Euro). The Jatts who had probably more Baloch affinity got sucked into the West Asian-heavy area.

I did the autosomal transfer for myself and got 99.2% South Asian and .8% Oceanic(Aborigine power ftw)

I STRONGLY suspect that they were just lazy and included mainly south Indians in their dataset(except tribals, leading to me getting extra Oceanian) because of the enormous non South Asian percentages that other groups are getting. In some cases it seems the west asian+Euro combines all or most of BMAC+MLBA steppe, which is insane.

Sapporo
02-05-2020, 09:21 AM
People who use that surname belong to either Bahun, Chhetri, or Dalit (Sarki) caste. But this is just my internet knowledge. I actually haven't met a single person who uses that surname ever in my life. Maybe poi bro knows more about it. Could that person belong to Dalit caste (going by low European %)?

Their South Asian % is even less than most NW pops (66-67%) and their West Asian is 23% so no way are they a dalit. It's possible they are recently part West Asian though?

agent_lime
02-05-2020, 09:23 AM
Their South Asian % is even less than most NW pops (66-67%) and their West Asian is 23% so no way are they a dalit. It's possible they are recently part West Asian though?

They are my distant relatives (2 of them). They are probably Aroras.

Sapporo
02-05-2020, 10:47 AM
They are my distant relatives (2 of them). They are probably Aroras.

Bashyal is used by Arora?

agent_lime
02-05-2020, 11:10 AM
Bashyal is used by Arora?

I haven't seen Bashyal as Arora. Just saying they are on my list.

agent_lime
02-05-2020, 12:18 PM
https://www.youtube.com/watch?v=kRZraaXPSNQ

Censored
02-05-2020, 12:40 PM
https://www.youtube.com/watch?v=kRZraaXPSNQ

11% Euro. This is EXACTLY the kind of crap that I worry about. People need to dump this company ASAP it’s misleading people.

Also, I cringe every time I hear her say “where I’m actually from”, your parents are from Pakistan, that’s your race.

agent_lime
02-05-2020, 01:10 PM
11% Euro. This is EXACTLY the kind of crap that I worry about. People need to dump this company ASAP it’s misleading people.

Also, I cringe every time I hear her say “where I’m actually from”, your parents are from Pakistan, that’s your race.

I think I am around 66% SA on this test. But "Kaur Beauty" is 57.9 %. MyHeritage is confusing people. They need to have recent and ancient ancestry not this weird hybrid.


https://www.youtube.com/watch?v=hGzwFAUiet8

Amber29
02-05-2020, 02:49 PM
I think I am around 66% SA on this test. But "Kaur Beauty" is 57.9 %. MyHeritage is confusing people. They need to have recent and ancient ancestry not this weird hybrid.


https://www.youtube.com/watch?v=hGzwFAUiet8

I remembe a girls video they said she was half chinese and half other asian when in fact she was full she retook the test etc and was told full south asian she was very pretty aswell

pnb123
02-05-2020, 03:02 PM
Their South Asian % is even less than most NW pops (66-67%) and their West Asian is 23% so no way are they a dalit. It's possible they are recently part West Asian though?
Idk what they could be. You have to ask them to find out.

Skyfall
02-05-2020, 04:25 PM
I think I am around 66% SA on this test. But "Kaur Beauty" is 57.9 %. MyHeritage is confusing people. They need to have recent and ancient ancestry not this weird hybrid.


https://www.youtube.com/watch?v=hGzwFAUiet8

:biggrin1: That tells me all I need to know about the "accuracy" of this test.

agent_lime
02-05-2020, 04:29 PM
:biggrin1: That tells me all I need to know about the "accuracy" of this test.

This Kaur chick is British so there is a chance she is mixed. Other than that Amber also hangs out around 70% if I remember correctly. Remember phenotype is not genotype. This person could be a Khatri or Jatt.

Skyfall
02-05-2020, 04:31 PM
This Kaur chick is British so there is a chance she is mixed. Other than that Amber also hangs out around 70% if I remember correctly. Remember phenotype is not genotype. This person could be a Khatri or Jatt.

Certainly plausible. But speculation aside, there is a high chance that these results are horribly wrong; a Hindu Khatri like yourself should be lower SA than her if she is a mixed caste (which she most likely is, owing to her surname).

agent_lime
02-05-2020, 04:40 PM
Certainly plausible. But speculation aside, there is a high chance that these results are horribly wrong; a Hindu Khatri like yourself should be lower SA than her if she is a mixed caste (which she most likely is, owing to her surname).

Not necessarily. There are plenty of Kaur Khatris and Jatts around. I know 2 Kaur Khatris. She could easily be 24-25% SI on Harappa. More likely is that one-two grandparents might that might be Sikhs from near Afghanistan Pakistan border or she inherited less AASI naturally.

poi
02-05-2020, 05:13 PM
People who use that surname belong to either Bahun, Chhetri, or Dalit (Sarki) caste. But this is just my internet knowledge. I actually haven't met a single person who uses that surname ever in my life. Maybe poi bro knows more about it. Could that person belong to Dalit caste (going by low European %)?


Their South Asian % is even less than most NW pops (66-67%) and their West Asian is 23% so no way are they a dalit. It's possible they are recently part West Asian though?

Looking at the Dr's spreadsheet, looks like random fluctuation. Just look at the parent with double WestAsia and the child with higher Euro.

https://i.imgur.com/CpR2zI1r.png

I have absolutely no knowledge of myheritagedna's breakdown (only know 23andme, ftdna, and ancestry), but I would not trust these companies' exact geographic breakdown, but they could be informative when you cross-compare within the same ethnic group or across the neighboring groups.

As far as the surname Bashyal, it is most likely one of Nepali Khas surnames.

Edit - looks like MyHeritageDNA's breakdown is so random even within the same ethnic group. @Dr, have you found something consistent/useful? For example, Harappa's breakdowns -- despite outdated components -- are consistent for the most part. And 23andme's recent ones are at least consistent within the same group. But MyHeritageDNA's breakdown looks almost useless.

prashantvaidwan
02-05-2020, 05:29 PM
Certainly plausible. But speculation aside, there is a high chance that these results are horribly wrong; a Hindu Khatri like yourself should be lower SA than her if she is a mixed caste (which she most likely is, owing to her surname).
Skyfall, I am really curious to know your ethnicity? Punjabi khatri or arora? Can you post ur harappa results if you got yourself tested?

Skyfall
02-05-2020, 05:30 PM
Sent via PM.

pegasus
02-05-2020, 06:15 PM
I think I am around 66% SA on this test. But "Kaur Beauty" is 57.9 %. MyHeritage is confusing people. They need to have recent and ancient ancestry not this weird hybrid.


https://www.youtube.com/watch?v=hGzwFAUiet8

If she is Jat , the odds of her having any recent modern European ancestry are practically none. If as a medical student she legit thinks she has 11% British ancestry or whatever, she needs to reconsider a new career.

Censored
02-05-2020, 06:40 PM
If she is Jat , the odds of her having any recent modern European ancestry are practically none. If as a medical student she legit thinks she has 11% British ancestry or whatever, she needs to reconsider a new career.

These sorts of beliefs are VERY common among laymen, trust me.

Dr_McNinja
02-05-2020, 06:50 PM
Looking at the Dr's spreadsheet, looks like random fluctuation. Just look at the parent with double WestAsia and the child with higher Euro.

https://i.imgur.com/CpR2zI1r.png

I have absolutely no knowledge of myheritagedna's breakdown (only know 23andme, ftdna, and ancestry), but I would not trust these companies' exact geographic breakdown, but they could be informative when you cross-compare within the same ethnic group or across the neighboring groups.

As far as the surname Bashyal, it is most likely one of Nepali Khas surnames.

Edit - looks like MyHeritageDNA's breakdown is so random even within the same ethnic group. @Dr, have you found something consistent/useful? For example, Harappa's breakdowns -- despite outdated components -- are consistent for the most part. And 23andme's recent ones are at least consistent within the same group. But MyHeritageDNA's breakdown looks almost useless.

Lol it is absolutely not useful, just entertaining. That's why I made the spreadsheet, so people could compare their results to others to make sense of it.

I wouldn't call it "wrong" though. It's population modeling and for the most part you can justify its conclusions, but it's all pointless because we have no idea how it works or what populations they've used. They don't even have paternal/maternal lineages there.

However, their user interface and family tree stuff is great and very easy to use and they have a sizable database of matches. I love that you can upload multiple kits for one person so you don't match 4 kits from the same person (unless they've uploaded them separately because they didn't know about that feature). I can't say enough great things about their website and mobile app. It blows 23andMe out of the water. FTDNA looks clunky/cumbersome in comparison too.

The fact it's not 100% South Asian for everyone means there's enough variance in the results that it can be made somewhat useful with a spreadsheet of other matches to compare yourself to. Don't want to leave MH users out in the cold. How is the overlap of MH raw data with gedmatch? Is it now better than FTDNA/LivingDNA?

Dr_McNinja
02-05-2020, 06:55 PM
I did the autosomal transfer for myself and got 99.2% South Asian and .8% Oceanic(Aborigine power ftw)

I STRONGLY suspect that they were just lazy and included mainly south Indians in their dataset(except tribals, leading to me getting extra Oceanian) because of the enormous non South Asian percentages that other groups are getting. In some cases it seems the west asian+Euro combines all or most of BMAC+MLBA steppe, which is insane.

Think of it this way: It's because South Asian and West Asian and even Central Asian all contain BMAC+Steppe. So whichever of those three you have a stronger affinity to, it'll pull most of your admixture into it. People from different regions/groups get different admixture "profiles" because of this. So people who have a South Asian draw will have their own pattern, and people with enough West Asian draw to compete with that will have their own.

This seems like a problem maybe limited to South Asians as far as I can tell. The moment you get past 40-50% South Asian, it resembles old 23andMe or FTDNA.

Censored
02-05-2020, 07:03 PM
Think of it this way: It's because South Asian and West Asian and even Central Asian all contain BMAC+Steppe. So whichever of those three you have a stronger affinity to, it'll pull most of your admixture into it. People from different regions/groups get different admixture "profiles" because of this. So people who have a South Asian draw will have their own pattern, and people with enough West Asian draw to compete with that will have their own.

This seems like a problem maybe limited to South Asians as far as I can tell. The moment you get past 40-50% South Asian, it resembles old 23andMe or FTDNA.

No, I don’t think any west Asians have BMAC besides POSSIBLY some in Persians. It’s a very central and secondarily south Asian phenomenon. Yeah there’s a bit of Steppe MLBA in West Asia especially among Iranians but nowhere near enough to justify the massive pull we see in some. We have been so diverged from the near East that it’s shocking. If anything central Asian or Euro should be higher.

Dr_McNinja
02-05-2020, 07:29 PM
No, I don’t think any west Asians have BMAC besides POSSIBLY some in Persians. It’s a very central and secondarily south Asian phenomenon. Yeah there’s a bit of Steppe MLBA in West Asia especially among Iranians but nowhere near enough to justify the massive pull we see in some. We have been so diverged from the near East that it’s shocking. If anything central Asian or Euro should be higher.

No, not in the region, in the components MyHeritage has. They're using multiple populations lumped together. So a little bit of BMAC ancestry in one is enough. I don't even mean BMAC specifically because they're such a recent composite population, but the usual ANE-ENF type (CHG/Iran_N) with Steppe/ANF mixed together. That signal is present in both West Asia and South Asia components in the MH test by way of a bunch of populations spanning from Turkey/Caucasus to Central Asia (according to their map). Central Asia component has it too though that also has East Eurasian.

Sapporo
02-06-2020, 02:51 AM
This Kaur chick is British so there is a chance she is mixed. Other than that Amber also hangs out around 70% if I remember correctly. Remember phenotype is not genotype. This person could be a Khatri or Jatt.

She's not part British. I've seen pictures and videos of her family before (many of my female relatives follow/know of her). AFAIK, she's Jatt Sikh (she's from the Bay Area like myself and we're the super majority of Sikhs here) but perhaps one that is heavily shifted toward Haryana/West UP Jaats based on her MyHeritage (it's very possible she scores like aaronbee or well within Jatt Sikh variation instead though). Also, as others have mentioned, MyHeritage is definitely not consistent for any South Asian ethnic/ethno-religious groups.


If she is Jat , the odds of her having any recent modern European ancestry are practically none. If as a medical student she legit thinks she has 11% British ancestry or whatever, she needs to reconsider a new career.

Giving her the benefit of the doubt, the average person (whether MD/PHD level educated or not) is highly ignorant of population genetics. I've tried explaining IVCp, BMAC, AASI and MLBA Steppe to my mama and massi before and they have a very misinformed view of Indo-Aryans or South Asian archaeology/migrations. They're all educated as well.


How is the overlap of MH raw data with gedmatch? Is it now better than FTDNA/LivingDNA?

The OG MyHeritage is equivalent to Ancestry DNA and 23andMe V3. The most current chip is comparable to the new FTDNA/23andMe V5. Living DNA's chip/raw data is probably the worst of the bunch in terms of overlap.

Censored
02-06-2020, 03:40 AM
Giving her the benefit of the doubt, the average person (whether MD/PHD level educated or not) is highly ignorant of population genetics. I've tried explaining IVCp, BMAC, AASI and MLBA Steppe to my mama and massi before and they have a very misinformed view of Indo-Aryans or South Asian archaeology/migrations. They're all educated as well.


Being educated is definitely not a bulwark against holding incorrect beliefs. In some cases it may actually cause one to have more due to incorrect assumptions made while learning. There's a lot of things that may seem logical on paper but are not once you dive deep and see what the research says.

For example we know that the northwest section of India has over the centuries been under many empires such as the Scythians, Greeks, Mongols(briefly), Persians, Arabs(briefly), and various Turkic empires like the Mughals. Given that, it would be logical to assume that the people living there have been genetically impacted by all these groups and to no surprise it's common for people, who are knowledgeable about history, to say that "north indians are very mixed", referring to admixture with these groups.

What people don't realize is that the ancestral alignment of the South Asian region was already complete by the Iron age and that whatever invasions happened afterward left little to no impact for the most part. And what impact from settlement did occur would largely be concentrated among urban Muslims like misanthropy or Khana(part Uzbek) rather than endogamous Hindus and Sikhs.

But this sort of misunderstanding is widespread and not limited to one region. A lot of people myself included at one point, would think that Spain has significant North African/Arab admixture after 7 centuries of Muslim rule but this seems to largely not be the case, and Spaniards are not much different from other southern Euros.

Dr_McNinja
02-08-2020, 12:33 AM
She's not part British. I've seen pictures and videos of her family before (many of my female relatives follow/know of her). AFAIK, she's Jatt Sikh (she's from the Bay Area like myself and we're the super majority of Sikhs here) but perhaps one that is heavily shifted toward Haryana/West UP Jaats based on her MyHeritage (it's very possible she scores like aaronbee or well within Jatt Sikh variation instead though). Also, as others have mentioned, MyHeritage is definitely not consistent for any South Asian ethnic/ethno-religious groups.



Giving her the benefit of the doubt, the average person (whether MD/PHD level educated or not) is highly ignorant of population genetics. I've tried explaining IVCp, BMAC, AASI and MLBA Steppe to my mama and massi before and they have a very misinformed view of Indo-Aryans or South Asian archaeology/migrations. They're all educated as well.



The OG MyHeritage is equivalent to Ancestry DNA and 23andMe V3. The most current chip is comparable to the new FTDNA/23andMe V5. Living DNA's chip/raw data is probably the worst of the bunch in terms of overlap.

Doesn't LivingDNA have more overlap with the Gedmatch calculators than V5 though? I haven't seen a new FTDNA kit on Gedmatch, does anyone have one?

Sapporo
02-08-2020, 05:39 AM
Doesn't LivingDNA have more overlap with the Gedmatch calculators than V5 though? I haven't seen a new FTDNA kit on Gedmatch, does anyone have one?

I haven't checked the exact amount of SNP overlap but from what I've noticed, 23andMe V5 GEDMatch results seems to match up against 23andMe V3 GEDMatch results better than Living DNA does.

Dr_McNinja
02-08-2020, 10:50 PM
I haven't checked the exact amount of SNP overlap but from what I've noticed, 23andMe V5 GEDMatch results seems to match up against 23andMe V3 GEDMatch results better than Living DNA does.

Just compared LivingDNA and V5 of same person on Gedmatch. Overlap with HarappaWorld:

LivingDNA: 132200 SNPs used in this evaluation

23andMe V5: 52590 SNPs used in this evaluation

Test was from Dec 2019 or thereabouts

Dr_McNinja
02-08-2020, 11:01 PM
I used my Dante results to make a V5 kit with WGSExtract. I actually tested on LivingDNA and on 23andMe V3:

V3:

Population
S-Indian 37.07 Pct
Baloch 34.85 Pct
Caucasian 10.75 Pct
NE-Euro 9.51 Pct
SE-Asian 0.12 Pct
Siberian 1.55 Pct
NE-Asian -
Papuan 0.85 Pct
American 1.05 Pct
Beringian 1.78 Pct
Mediterranean 1.24 Pct
SW-Asian 1.23 Pct
San -
E-African -
Pygmy -
W-African -

V5:

Population
S-Indian 35.95 Pct
Baloch 35.13 Pct
Caucasian 11.44 Pct
NE-Euro 10.52 Pct
SE-Asian -
Siberian 1.03 Pct
NE-Asian 0.89 Pct
Papuan 1.22 Pct
American 1.71 Pct
Beringian 1.35 Pct
Mediterranean -
SW-Asian 0.41 Pct
San -
E-African -
Pygmy 0.35 Pct
W-African -

LivingDNA:

Population
S-Indian 37.24 Pct
Baloch 34.30 Pct
Caucasian 10.37 Pct
NE-Euro 9.34 Pct
SE-Asian -
Siberian 1.61 Pct
NE-Asian 0.22 Pct
Papuan 0.94 Pct
American 0.88 Pct
Beringian 1.74 Pct
Mediterranean 1.31 Pct
SW-Asian 1.84 Pct
San -
E-African -
Pygmy 0.22 Pct
W-African -

Dr_McNinja
02-08-2020, 11:05 PM
For comparison:

Dante WGS 30X (186764 SNPs):

Population
S-Indian 36.91 Pct
Baloch 34.68 Pct
Caucasian 10.91 Pct
NE-Euro 9.51 Pct
SE-Asian 0.12 Pct
Siberian 1.62 Pct
NE-Asian -
Papuan 0.90 Pct
American 1.03 Pct
Beringian 1.76 Pct
Mediterranean 1.29 Pct
SW-Asian 1.26 Pct
San -
E-African -
Pygmy -
W-African -

V3+FTDNAFF+LivingDNA+Geno2 (185316 SNPs):

Population
S-Indian 36.96 Pct
Baloch 34.78 Pct
Caucasian 10.66 Pct
NE-Euro 9.62 Pct
SE-Asian 0.18 Pct
Siberian 1.61 Pct
NE-Asian -
Papuan 0.87 Pct
American 1.00 Pct
Beringian 1.74 Pct
Mediterranean 1.30 Pct
SW-Asian 1.27 Pct
San -
E-African -
Pygmy -
W-African -

EDIT:

There were always a few SNPs left over because every kit, even Dante, had some missing SNPs. To get every last SNP, I combined all my kits and ran DIYDodecadWin. It was picking up a couple SNPs from each kit to complete the whole thing.

That's 23andMe V3, FTDNA FF (old one 2016-2019), LivingDNA (12/2019), Geno 2.0 (old one with ~150k SNPs), Full Genomes Corp (I had a ~250k SNP file of autosomal data left over from my Y Elite/Prime order), Dante Labs WGS 30X (the 'Combined Kits' and '23andMe API SNPs' options in WGSExtractBeta). Total is 2,156,232 SNPs.

So, this is 100% genotyping rate, zero SNPs missing in DIYDodecadWin:

36.79% S-Indian
34.80% Baloch
10.59% Caucasian
9.49% NE-Euro
0.18% SE-Asian
1.46% Siberian
0.02% NE-Asian
0.93% Papuan
0.99% American
1.77% Beringian
1.47% Mediterranean
1.53% SW-Asian
0.00% San
0.00% E-African
0.00% Pygmy
0.00% W-African

4-Ancestors Oracle:

Using 4 populations approximation:
1 north-kannadi+punjabi-jatt-sikh+punjabi-jatt-sikh+pashtun-uthmankhel @ 1.181724
2 hakkipikki+haryana-jatt+punjabi-jatt-muslim+nuristani @ 1.183359
3 bhil+punjabi-jatt-muslim+punjabi-jatt-sikh+pashtun-uthmankhel @ 1.188362
4 chamar+punjabi-jatt-muslim+punjabi-jatt-sikh+pashtun-uthmankhel @ 1.202325
5 hakkipikki+haryana-jatt+kalash+punjabi-jatt-muslim @ 1.210486

It's not being tokenized on gedmatch though, might be too big.

EDIT again: Got tokenized but one randomly loses a bunch of SNPs after tokenization. I uploaded another attempt which got through, mostly. That one's Gedmatch Harappa:

(186768 SNPs)

Population
S-Indian 36.85 Pct
Baloch 34.70 Pct
Caucasian 10.69 Pct
NE-Euro 9.63 Pct
SE-Asian 0.16 Pct
Siberian 1.65 Pct
NE-Asian -
Papuan 0.87 Pct
American 1.02 Pct
Beringian 1.74 Pct
Mediterranean 1.36 Pct
SW-Asian 1.33 Pct
San -
E-African -
Pygmy -
W-African -

Using 4 populations approximation:
++++++++++++++++++++++++++++++++++++++++++++++++++ ++++++++++++++++++++++++++++++++++++++++++++++++++ ++++++++++++++++++++++++
1 haryana-jatt_harappa + kalash_hgdp + north-kannadi_chaubey + punjabi-jatt-muslim_harappa @ 1.123203
2 hakkipikki_metspalu + haryana-jatt_harappa + kalash_hgdp + punjabi-jatt-muslim_harappa @ 1.166674
3 haryana-jatt_harappa + kalash_hgdp + kurumba_reich + punjabi-khatri_harappa @ 1.176687
4 hakkipikki_metspalu + haryana-jatt_harappa + kalash_hgdp + punjabi-khatri_harappa @ 1.178176
5 hakkipikki_metspalu + haryana-jatt_harappa + kalash_hgdp + pushtikar-brahmin_harappa @ 1.218471
6 haryana-jatt_harappa + kalash_hgdp + kurumba_reich + punjabi-jatt-muslim_harappa @ 1.305389
7 haryana-jatt_harappa + north-kannadi_chaubey + pathan_hgdp + punjabi-khatri_harappa @ 1.308008

I think it's something with LivingDNA's raw data. Davidski also said their raw data was weird and giving him problems, and people had to use Excel to fix it. Not sure what it is.

subzero85
07-08-2020, 11:02 PM
38370

Mine is based on a 23andMe transfer.

subzero85
07-08-2020, 11:03 PM
38370

Mine is based on a 23andMe transfer.

85% South Asian

5% Central Asian
5% West Asian
5% Balkan

Dr_McNinja
07-14-2020, 06:38 PM
MyHeritage tests on the GSA chip similar to FTDNA but there's small differences. Nonetheless, a transfer from Family Finder V3 (after March-April of 2019) gives the closest results to MyHeritage's own chip (and their stuff is sequenced by FTDNA's lab anyway).

For a comparison of autosomal testing companies, here are a few of mine:


DanteLabs-9999,36.93,34.68,10.90,9.49,0.09,1.67,0.01,0.89,1. 07,1.65,1.29,1.34,0,0,0,0
MyHeritageV1-OldChip,36.95,34.78,10.80,9.47,0.09,1.61,0.01,0.90 ,1.12,1.71,1.26,1.30,0,0,0,0
FTDNAV3+LivingDNA+FTDNAV3Imputed,36.78,34.51,10.79 ,9.64,0.32,1.48,0.02,0.79,1.21,1.53,1.29,1.37,0.00 ,0.00,0.12,0.15
FTDNAV3+LivingDNA,37.11,34.31,11.20,9.29,0.02,1.60 ,0.03,1.08,1.08,1.63,1.17,1.32,0.00,0.00,0.17,0.00
FTDNAV3Imputed,36.79,34.48,10.68,9.63,0.39,1.54,0. 02,0.79,1.33,1.62,1.12,1.30,0,0,0.14,0.19
FTDNAV3,36.65,35.13,12.62,9.1,0,1.6,0,0.91,1.45,1. 56,0.35,0.25,0,0.07,0.31,0
LivingDNA-Current,37.32,34.32,10.41,9.16,0.04,1.61,0.10,0.99 ,0.96,1.73,1.43,1.71,0,0,0.22,0
23andMeV3,37.09,34.88,10.77,9.51,0.10,1.62,0.01,0. 84,1.01,1.71,1.24,1.21,0,0,0,0
FTDNAV2-OldChip,37.09,34.88,10.77,9.51,0.10,1.62,0.01,0.84 ,1.01,1.71,1.24,1.21,0,0,0,0
23andMeV5,35.97,35.16,11.42,10.47,0.01,1.01,0.87,1 .23,1.71,1.35,0.02,0.42,0,0.01,0.34,0
23andMeV4,36.89,34.97,10.84,9.41,0.27,1.57,0.02,0. 84,0.82,1.79,1.12,1.45,0,0,0,0


Distance to: DanteLabs-9999
0.18193405 MyHeritageV1-OldChip
0.33852622 23andMeV3
0.33852622 FTDNAV2-OldChip
0.50685304 FTDNAV3+LivingDNA+FTDNAV3Imputed
0.51361464 23andMeV4
0.62433965 FTDNAV3+LivingDNA
0.63190189 FTDNAV3Imputed
0.93717661 LivingDNA-Current
2.39376273 FTDNAV3
2.59896133 23andMeV5

Distance to: 23andMeV3
0.00000000 FTDNAV2-OldChip
0.23769729 MyHeritageV1-OldChip
0.33852622 DanteLabs-9999
0.45716518 23andMeV4
0.67697858 FTDNAV3+LivingDNA+FTDNAV3Imputed
0.74451326 FTDNAV3Imputed
0.82498485 FTDNAV3+LivingDNA
0.99513818 LivingDNA-Current
2.42752961 FTDNAV3
2.60702896 23andMeV5FTDNAV3 is interchangeable with MyHeritageV2 (current MyHeritage testing platform) in that list.

It appears the best way to get a decent overlap with Gedmatch calculators today is to combine kits from either Family Tree DNA or MyHeritage with LivingDNA and then with an imputed VCF of the former data (FTDNA or MH) from DNA.Land (doesn't accept LivingDNA at the moment). Without the imputed data that's around 75%.

Dr_McNinja
07-29-2020, 03:00 PM
So if you're uploading your raw data from another platform to MyHeritage, you want to upload it in Family Tree DNA Family Finder V3 format so your results would match what you would get if you took MyHeritage's own DNA test. That FTDNA platform is called V1.1 or V2 in DNA Kit Studio 2.8 (whichever is the newest one in the selection). You can use your autosomal data from a WGS test or an imputed VCF from a site like DNA.Land in DNA Kit Studio to make the compatible file (WGS Extract Beta should allow you to make the V3 format autosomal file directly from the BAM, avoiding the need for DNA Kit Studio). You might have to tinker with the Sort by Position vs. Sort by RSID settings to get MyHeritage to take it. I even pasted some empty fields from the beginning of an actual Family Finder V3 raw data file into the beginning of the DNA Kit Studio generated file to make sure it would be accepted. Not sure if that made a difference or not. But I finally did get it to accept a FF V3 format file made using a DNA.Land imputed VCF of a 23andMe V5 test.

I think you could just upload that directly to FTDNA for autosomal transfer as well (might recognize it as MyHeritage?), but not sure anyone gives a damn about their current version of myOrigins, lol. I guess if v3.0 is popular?

Also, to add to the previous post: Ancestry DNA is still using the old Illumina Express type chip I think and has very good overlap with Gedmatch and old calculators. It's a saliva test like 23andMe's. If you need to take a test, that should be the #1 choice if the price isn't an issue (the other companies drop their prices to as low as $50 often). LivingDNA is next with 60-70% overlap. Then the FTDNA/MH tests at ~38-40% and then last 23andMe V5 at ~28-30%.

MonkeyDLuffy
07-29-2020, 03:58 PM
So if you're uploading your raw data from another platform to MyHeritage, you want to upload it in Family Tree DNA Family Finder V3 format so your results would match what you would get if you took MyHeritage's own DNA test. That FTDNA platform is called V1.1 or V2 in DNA Kit Studio 2.8 (whichever is the newest one in the selection). You can use your autosomal data from a WGS test or an imputed VCF from a site like DNA.Land in DNA Kit Studio to make the compatible file (WGS Extract Beta should allow you to make the V3 format autosomal file directly from the BAM, avoiding the need for DNA Kit Studio). You might have to tinker with the Sort by Position vs. Sort by RSID settings to get MyHeritage to take it. I even pasted some empty fields from the beginning of an actual Family Finder V3 raw data file into the beginning of the DNA Kit Studio generated file to make sure it would be accepted. Not sure if that made a difference or not. But I finally did get it to accept a FF V3 format file made using a DNA.Land imputed VCF of a 23andMe V5 test.

I think you could just upload that directly to FTDNA for autosomal transfer as well (might recognize it as MyHeritage?), but not sure anyone gives a damn about their current version of myOrigins, lol. I guess if v3.0 is popular?

Also, to add to the previous post: Ancestry DNA is still using the old Illumina Express type chip I think and has very good overlap with Gedmatch and old calculators. It's a saliva test like 23andMe's. If you need to take a test, that should be the #1 choice if the price isn't an issue (the other companies drop their prices to as low as $50 often). LivingDNA is next with 60-70% overlap. Then the FTDNA/MH tests at ~38-40% and then last 23andMe V5 at ~28-30%.

My heritage ancestry feature used to be free for transfer, now they charge money for ancestry composition.

Dr_McNinja
07-29-2020, 04:30 PM
My heritage ancestry feature used to be free for transfer, now they charge money for ancestry composition.

Oh yeah, then probably not worth it, might as well get their DNA test. I subscribe to the site through one of their plans so I continue to get ethnicity estimate results for all transfers. I re-uploaded some of my relatives using this method just to see how it works (originally a combined kit which should have had enough data for FTDNAV3, and then later just imputed kits from 23andMe V4/V5). I guess some people may really want to do a transfer to MH and pay for it, so this may be helpful for them.

Amber29
12-23-2020, 11:33 PM
updated results:

41941

subzero85
12-23-2020, 11:37 PM
41947

Sapporo
12-24-2020, 01:54 AM
My updated results:

23andMe V3 Transfer:

41950

My AncestryDNA transfer is basically the same.

Official MyHeritage (one of the newer chips):

41951

My guess is the Punjabi samples are a mix of PJL? Otherwise, not sure why we would still get West Asia and Europe.


updated results:

41941

Do your results show regional breakdown for South Asia? Like Punjab or AJK?

Jatt1
12-24-2020, 03:06 AM
My updated results:

23andMe V3 Transfer:

41950

My AncestryDNA transfer is basically the same.

Official MyHeritage (newest chip):

41951

My guess is the Punjabi samples are a mix of PJL? Otherwise, not sure why we would still get West Asia and Europe.



Do your results show regional breakdown for South Asia? Like Punjab or AJK?

MyHeritage shows West Asia and Europe on mine too, mine is from old chip though. I don't know how to post it here.

MonkeyDLuffy
12-24-2020, 03:34 AM
Mine:

41952

Jatt1
12-24-2020, 04:00 AM
Mine:

Asia
South Asian
65.8%

Pakistan (Punjab) and India
India (Punjab) and Pakistan (Punjab)
India

West Asian
20.6%

Central Asian
4.2%

Europe
Finnish
4.6%

Baltic
2.6%

Irish, Scottish, and Welsh
1.3%

Scandinavian
0.9%

Show all available regions

agent_lime
12-24-2020, 07:06 AM
Here are my updated results. I am not getting a Punjabi, here are my 4 genetic groups.

Indians and Pakistanis in India and in Pakistan, and some of their descendants in England Confidence level: High
Indians in India Confidence level: Low Confidence level: Low
India, Pakistan, Kenya, Tanzania, Saudi Arabia and Kuwait Confidence level: Low
Welsh in Wales (Carmarthenshire and Swansea) and some in England Confidence level: Low (LOL)

https://i.imgur.com/Vqh3CNV.jpg

https://i.imgur.com/qJxm6gj.jpg

Amber29
12-24-2020, 09:00 AM
My updated results:

23andMe V3 Transfer:

41950

My AncestryDNA transfer is basically the same.

Official MyHeritage (one of the newer chips):

41951

My guess is the Punjabi samples are a mix of PJL? Otherwise, not sure why we would still get West Asia and Europe.



Do your results show regional breakdown for South Asia? Like Punjab or AJK?

I shall post it now

41977

My dads

41978


My cousins Merry

41979

Im showint different peeops due to the different out looks for comparisons... i need to unlock other family members so cant see some members yet to see other differences lol

Ahmed Ali
12-24-2020, 11:03 AM
My grandad's updated:

South Asian 81.5% (Pakistan Punjab and India) (Afghanistan)
West Asian 1.2%
Irish, Scottish and Welsh 11.1%
Finnish 2.6%
Italian 2.3%
Mesoamerican and American 1.3%

(sorry for the lack of screenshot - having tech issues these days)

His results seem to tally with other platforms/calcs. Afghanistan at high confidence is interesting. West Asian also surprisingly low!

agent_lime
12-24-2020, 01:41 PM
My grandad's updated:

South Asian 81.5% (Pakistan Punjab and India) (Afghanistan)
West Asian 1.2%
Irish, Scottish and Welsh 11.1%
Finnish 2.6%
Italian 2.3%
Mesoamerican and American 1.3%

(sorry for the lack of screenshot - having tech issues these days)

His results seem to tally with other platforms/calcs. Afghanistan at high confidence is interesting. West Asian also surprisingly low!

Can you list your genetic groups?

Ahmed Ali
12-24-2020, 02:01 PM
Can you list your genetic groups?

Sure. The two genetic groups my grandad gets are "Pakistan (Punjab) and India" and "Afghanistan" (both high confidence).

When I click on them, the heatmap seems to spike around Sialkot (for Punjab) and Kabul and Ghazni for Afghanistan.. not sure if those really mean anything though..

agent_lime
12-24-2020, 02:15 PM
Sure. The two genetic groups my grandad gets are "Pakistan (Punjab) and India" and "Afghanistan" (both high confidence).

When I click on them, the heatmap seems to spike around Sialkot (for Punjab) and Kabul and Ghazni for Afghanistan.. not sure if those really mean anything though..

Thanks Ahmed. It makes sense since you have some real Afghan ancestry. I just wanted to see if you were getting odd genetic groups like everyone else. Like Amber's not that much different from you but on Heritage she seems quite different.

Ahmed Ali
12-24-2020, 02:55 PM
Thanks Ahmed. It makes sense since you have some real Afghan ancestry. I just wanted to see if you were getting odd genetic groups like everyone else. Like Amber's not that much different from you but on Heritage she seems quite different.

I agree. There seems to be a disparity in the way MyHeritage is construing the results of NW South Asians.. It seemed to do a pretty good job for my grandpa (even picking up the minor Italian/Iberian component he gets elsewhere) but it appears to apportion Steppe/Neolithic Iranian ancestry quite differently for different people.. E.g. I saw some users are getting central asian, others Baltic and Eastern euro, my grandad got Mesoamerican etc.. I'm assuming these are all attempts to capture our steppic admix..

on a related note, any thoughts on what "Irish, Scottish and Welsh" is proxying for? I see that most of us are getting in but in different amounts..

tipirneni
12-24-2020, 03:12 PM
I agree. There seems to be a disparity in the way MyHeritage is construing the results of NW South Asians.. It seemed to do a pretty good job for my grandpa (even picking up the minor Italian/Iberian component he gets elsewhere) but it appears to apportion Steppe/Neolithic Iranian ancestry quite differently for different people.. E.g. I saw some users are getting central asian, others Baltic and Eastern euro, my grandad got Mesoamerican etc.. I'm assuming these are all attempts to capture our steppic admix..

on a related note, any thoughts on what "Irish, Scottish and Welsh" is proxying for? I see that most of us are getting in but in different amounts..

There is some cluster of Irish,Scotissh and Welsh along with Scandinavian shows up on mine too more than 20% of the matches on segment matches bigger than 10cM. This might be normal for certain kind of foreign admixture

This is my match ethnicity list which shows that distribution. Some European might have multiple categories listed.
https://i.imgur.com/AtkGlM2.png

subzero85
12-24-2020, 03:42 PM
41947


Black-Sea Germans in Ukraine (Odessa) and in Russia is a genetic group of mine.

agent_lime
12-24-2020, 06:45 PM
I agree. There seems to be a disparity in the way MyHeritage is construing the results of NW South Asians.. It seemed to do a pretty good job for my grandpa (even picking up the minor Italian/Iberian component he gets elsewhere) but it appears to apportion Steppe/Neolithic Iranian ancestry quite differently for different people.. E.g. I saw some users are getting central asian, others Baltic and Eastern euro, my grandad got Mesoamerican etc.. I'm assuming these are all attempts to capture our steppic admix..

on a related note, any thoughts on what "Irish, Scottish and Welsh" is proxying for? I see that most of us are getting in but in different amounts..

I think My Heritage is picking up excess from it's base ancestries references. In the case of Punjabis, it might be using PJL. Steppe MLBA and North Euro is probably close enough to pick up as a proxy. That accounts for Scottish and Welsh. Plus there is reverse South Asian gene flow into England as well because of the Raj so that might be getting picked up a bit.

In my case the Indians and Pakistanis from England make sense because I have a bunch of English relatives with a few % South Asian. My maternal great grandfather was also a barrister in London for a few years, who knows what he was up to.

Ahmed Ali
12-24-2020, 09:55 PM
I think My Heritage is picking up excess from it's base ancestries references. In the case of Punjabis, it might be using PJL. Steppe MLBA and North Euro is probably close enough to pick up as a proxy. That accounts for Scottish and Welsh. Plus there is reverse South Asian gene flow into England as well because of the Raj so that might be getting picked up a bit.

In my case the Indians and Pakistanis from England make sense because I have a bunch of English relatives with a few % South Asian. My maternal great grandfather was also a barrister in London for a few years, who knows what he was up to.

That last sentence made me smile - i'm a barrister based in London too :)

ssamlal
12-24-2020, 11:22 PM
My results; mum and brother_VS also get the same results.

42014

Brother_RS

42015

Paternal Uncle

42016

Niece_TS

42017

Cousin_LS

42018

Strange that my family and I all get the "Flemish in Belgium (Flanders)" genetic group but not the one for T&T :(

AlluGobi
12-25-2020, 09:22 PM
My results

42042

Khan of Khura
12-26-2020, 03:17 PM
Hi,

I am related to AlluGobi, and these are my results :) 42063

karnalIroh
12-26-2020, 11:55 PM
i am really confused is this the update everyone is talking about
42084

agent_lime
12-27-2020, 04:58 AM
i am really confused is this the update everyone is talking about
42084

Yes. But it seems they have references for your ethnic group/ close relatives. So your is more reasonable than Khatris/ Jatts etc

karnalIroh
12-27-2020, 05:12 AM
Yes. But it seems they have references for your ethnic group/ close relatives. So your is more reasonable than Khatris/ Jatts etc
so by northwest england do they mean the mirpuris over there that would make sense but what of algeria and morroco?

agent_lime
12-27-2020, 05:20 AM
so by northwest england do they mean the mirpuris over there that would make sense but what of algeria and morroco?

My Heritage is just not very good. You have excess Steppe/ WANA vs the references they have, that's all this means, nothing recent. It's not picking up Mirpuris for NW England.

laltota
12-27-2020, 10:24 AM
My results

42042

Even though you are 1/4 Awan and 1/2 Kashmiri Butt, the only group shown for you under South Asian is "Indians in Suriname (Paramaribo) and in Guyana". It has this group also for Khan of Khura, as well as "Pakistan (Punjab) and India" and "India and Pakistan". It is strange that it does not mention any other groups for you apart from Suriname and Guyana.

AlluGobi
12-27-2020, 05:45 PM
Even though you are 1/4 Awan and 1/2 Kashmiri Butt, the only group shown for you under South Asian is "Indians in Suriname (Paramaribo) and in Guyana". It has this group also for Khan of Khura, as well as "Pakistan (Punjab) and India" and "India and Pakistan". It is strange that it does not mention any other groups for you apart from Suriname and Guyana.

Yeah, I am not sure why this is. Could it be because the raw files I transferred over was from 23andme V4? Khan of Khura is on V5. Or is just that MyHeritage is just not much accurate in general?

Also forgot to post my sisters' (Raw data from Ancestry):

42094

00simba
01-03-2021, 07:22 PM
Hi everyone. For some background both parents sides are Muslim from Amritsar area and moms side is of Mughal descent.
I've attached my results (first pic) and my moms. Also for some reason I didn't get any regional breakdowns. If anyone can answer why I'd appreciate it.

4226642267

a_kayani987
02-14-2021, 02:33 AM
my results im a gakkar my parents are from jhelum

https://imgur.com/nMLk7g5

misanthropy
02-15-2021, 05:42 AM
Should I convert my 23andme v5.9 to v3 before transferring? Any difference on this?

royaljoker
02-21-2021, 09:52 PM
What is generally the best of the three. From my FTDNA, My heritage and Living DNA.

I've got Scandinavian DNA on my heritage for a transferred result of a family member but non of the other two show that. Not sure what to make of it. Its just a bit under 10%.

Kapisa
02-21-2021, 10:33 PM
What is generally the best of the three. From my FTDNA, My heritage and Living DNA.

I've got Scandinavian DNA on my heritage for a transferred result of a family member but non of the other two show that. Not sure what to make of it. Its just a bit under 10%.

Recombination usually leads to family members getting different proportions of ancestral alleles. From what I have read FTDNA is better.

royaljoker
02-21-2021, 11:12 PM
Recombination usually leads to family members getting different proportions of ancestral alleles. From what I have read FTDNA is better.

Thanks, but also I meant that the same person does'nt show Scandinavian ancestry on FTDNA and Living DNA. Its only on My Heritage.

Also how good was the now defunct genographic's website by Nat Geo. We had the genographic results for a family member (using the Helix version of the kit) who's now passed away before we could transfer them. That person only showed two ethnic groups (about 90% South Asia and 10% Anatolia) but now HelixDNA is showing one more ethnic groups. The helix dna now show's 75.5% South Asia, 23.1% Middle East & North Africa (including Anatolia) and somewhat confusingly Beringian at 1.3%. I'd also be really really really grateful if someone knew how I could transfer the Helix version of Genographic to other websites.

Kapisa
02-21-2021, 11:38 PM
Thanks, but also I meant that the same person does'nt show Scandinavian ancestry on FTDNA and Living DNA. Its only on My Heritage.

Also how good was the now defunct genographic's website by Nat Geo. We had the genographic results for a family member (using the Helix version of the kit) who's now passed away before we could transfer them. That person only showed two ethnic groups (about 90% South Asia and 10% Anatolia) but now HelixDNA is showing one more ethnic groups. The helix dna now show's 75.5% South Asia, 23.1% Middle East & North Africa (including Anatolia) and somewhat confusingly Beringian at 1.3%. I'd also be really really really grateful if someone knew how I could transfer the Helix version of Genographic to other websites.

Every testing company has different ways of assigning alleles to different population clusters, different threshold. Best thing would be to look for consistancy, if excess European alleles are assigned to a certain European population cluster than see if 'similar' if not 'identical' cluster shows up in other results. I think Scandinavian cluster would not necessarily be the same as Mesopotamian or Anatolian since both populations peak in different ancestoral sources. I would trust FTDNA and LivingDNA over My Heritage.

royaljoker
02-21-2021, 11:43 PM
Every testing company has different ways of assigning alleles to different population clusters, different threshold. Best thing would be to look for consistancy, if excess European alleles are assigned to a certain European population cluster than see if 'similar' if not 'identical' cluster shows up in other results. I think Scandinavian cluster would not necessarily be the same as Mesopotamian or Anatolian since both populations peak in different ancestoral sources. I would trust FTDNA and LivingDNA over My Heritage.

Thanks a lot. This was really helpful

pakistani
03-01-2021, 06:15 PM
43622

Ummmm yeah, definitely doesn't seem that accurate haha. Is the Eastern European due to elevated Steppe admixture or something?

Azbuzz
03-05-2021, 03:18 PM
My Myheritage results-23 and me upload

59.8 percent West Asian from Iran
25.3 percent South Asian
8.6 percent Middle Eastern
6.3 percent Ashkenazi Jewish

Jatt1
03-05-2021, 11:17 PM
My Myheritage results-23 and me upload

59.8 percent West Asian from Iran
25.3 percent South Asian
8.6 percent Middle Eastern
6.3 percent Ashkenazi Jewish

Did you test with MyHeritage?

Azbuzz
03-06-2021, 04:46 AM
Did you test with MyHeritage?

No I tested with 23andme and then uploaded my 23andme data onto MyHeritage.