Page 3 of 3 FirstFirst 123
Results 21 to 28 of 28

Thread: Pashtun ancestry?

  1. #21
    Registered Users
    Posts
    99
    Sex

    Quote Originally Posted by Dr_McNinja View Post
    What's your background? Your Harappa results are unusual and appear mixed. Though if that's 23andMe V5, that's only ~30% overlap or so, right? Have you imputed your 23andMe raw data at DNA.Land when that still worked?
    I am an Urdu speaker on both sides. To be more precise, and only including what I've been able to sufficiently verify:

    My mother's side was based in and around Delhi for centuries (since at least the 1300s). They primarily spoke Urdu and Persian and some knew Arabic as a scholarly profession. Ancestors not from the subcontinent came from Bukhara in the 1300s, Hadhramaut in the 1500s, and Baghdad in the 1700s. Endogamy was practiced on this branch quite a fair amount. I shared a fairly detailed family tree of my mother's side here if it may interest anyone or yield anything, but I'm fairly sure there weren't any Pashtuns on that side.

    My father's side is where I'm much less certain. There's the topic of discussion Mr. Hashmatullah Khan who was one of my great-great grandfathers, then there's another UP Khan I mentioned here who also claimed to have Pashtun ancestry. My direct line was a UP Syed, and the last great-great grandfather I know nothing about was named "Hakeem Faizzal Hassan." As far as I know everyone spoke Urdu on this side too.

    So that's what I'm working with. Seems like Urdu speakers in general are the mutts of the subcontinent.

    I actually did do DNA.land about two years ago, attaching the results.dna.land.png
    Last edited by pakistani; 07-23-2020 at 07:54 PM.

  2. The Following 10 Users Say Thank You to pakistani For This Useful Post:

     Ahmed Ali (07-23-2020),  khanabadoshi (07-24-2020),  Kulin (07-24-2020),  laltota (07-23-2020),  MonkeyDLuffy (07-24-2020),  parasar (07-24-2020),  Reza (07-24-2020),  Sapporo (07-24-2020),  subzero85 (07-23-2020),  ThaYamamoto (07-23-2020)

  3. #22
    Registered Users
    Posts
    3,360
    Sex
    Location
    Chicago
    Ethnicity
    Baloch Kashmiri Uzbek Kho
    Nationality
    USA
    Y-DNA (P)
    R-Y17491 > R-YP4858
    mtDNA (M)
    A8a

    Pakistan United States of America
    You need a plan of action to tackle this. Firstly, you to need establish or estimate timeframes. Do they match up with historical events that make that migration plausible? Secondly, you need to understand the relative relevance of a distant ancestor to your genetic profile and the possibility of its detection.



    THEORHETICALLY you need to test:
    Great-great Grandparent to detect an unadmixed ancestor 10 generations ago
    Great Grandparent to detect an unadmixed ancestor 9 generations ago
    Grandparent to detect an unadmixed ancestor 8 generations ago
    Parent to detect an unadmixed ancestor 7 generations ago
    Yourself to detect an unadmixed ancestor 6 generations ago.

    PRACTICALLY, you need to test:
    Great-great Grandparent to detect an unadmixed ancestor 8 generations ago
    Great Grandparent to detect an unadmixed ancestor 7 generations ago
    Grandparent to detect an unadmixed ancestor 6 generations ago
    Parent to detect an unadmixed ancestor 5 generations ago
    Yourself to detect an unadmixed ancestor 4 generations ago.

    The easiest thing to detect would be Iraqi ancestry from the 1700s -- it is the most recent in time and a stream of ancestry completely off the S/SC Asian genetic cline. Even though the Yemeni linkage is too far removed to detect on its own, the Iraqi addition later would compound the MENA stream of ancestry as a whole; this coupled with the fact that Muslims tend to practice cousin-marriages means that your MENA ancestry should be detectable at the very least in your grandparent(s) (if we are talking mid- to late- 18th century Baghdadi ancestor).

    Teasing out ancestry along the S/SC Asian genetic cline itself is tedious and challenging -- this is mostly because you are already on said cline, and not on the opposite end of the cline from a Pashtun, but rather somewhere in the middle of it. To put this into perspective and temper expectations... my grandmother is 75% Kho/Chitrali, my mother 37.5%, and myself 18.75% -- forget me... my MOTHER can be modeled without Kho OR any of the stand-ins (like Tajik or Pashtun). Let's say I didn't know what my grandmother was, I never tested her or my mother, and all I had was my own result... I'd never in a million years suspect Kho ancestry or assume it. I might have suspected a general stream of ancestry from the Northwest and I'd likely have attributed it to Pashtuns since they are the major population, but more likely than not, I'd just assume I was a Jatt or Gujjar (especially if I knew nothing specific of any of my grandparent's ancestry). Now the caveat is that this is relative... if your grandmother has the same profile as mine, but you are sitting in Uttar Pradesh... this stream of ancestry would make you more distinct from the population around you than it distinguishes me from the population around me in Punjab, and thus, it would be far more useful. You could come to much stronger conclusions than I could; in fact, you'd probably have to suspect a recent ancestor from further NW in South Asia, at the very least. However, even you would likely never assume Chitrali or even Burusho, or Shina... you might settle on Jatt or Kashmiri, or Tajik or Pashtun.

    So what to do? Distinguish between what you know and suspect. Assume all stories are true and sit there and test this. How? Well, you seem to have a strong family tree game going on... so assume so-and-so is 100% something and work your way down. Assume all non-related wives are local to wherever that ancestor is in time unless otherwise specified. Assign everyone something and work your way down to you. You'll have a theoretical sense of what you "should" be if "all the stories are true". Then run a bunch of admix calculators and see if the fits are decent with those populations alone. Eliminate, add, rework the theory... repeat.

    It is hard work. I am trying to figure out what my maternal Grandfather's grandmother was... and as of yet I can't. My best guess is she is from either Shamli in Uttar Pradesh, Bharuch in Gujarat, or Maharashtrian... and that is from 23andme relatives being related to specific people in my family only descended from this woman. Admix calculators have been relatively useless here, except that they point me eastward.
    Last edited by khanabadoshi; 07-24-2020 at 03:21 AM.
    “Chahar chez est tohfay Multan, Gard-o- Garma, Gada-o- Goristan”.

    Four things are the gift of Multan: Dusty winds, hot seasons, beggars and graveyards.




  4. The Following 23 Users Say Thank You to khanabadoshi For This Useful Post:

     agent_lime (07-24-2020),  Ahmed Ali (07-24-2020),  Amber29 (07-24-2020),  Coldmountains (07-24-2020),  deuterium_1 (07-24-2020),  discreetmaverick (07-24-2020),  Dr_McNinja (07-24-2020),  GenoPlot (07-24-2020),  gudar (07-24-2020),  heksindhi (07-24-2020),  Kulin (07-24-2020),  misanthropy (07-25-2020),  MonkeyDLuffy (07-24-2020),  pakistani (07-24-2020),  parasar (07-26-2020),  pegasus (07-24-2020),  Raza94 (07-24-2020),  Reza (07-24-2020),  Riverman (07-24-2020),  Sapporo (07-24-2020),  subzero85 (07-24-2020),  ThaYamamoto (07-24-2020),  thejkhan (07-24-2020)

  5. #23
    Registered Users
    Posts
    2,622
    Sex
    Location
    New York
    Ethnicity
    South Asian
    Nationality
    American
    Y-DNA (P)
    J2b2-M241 (J-Y978*)
    mtDNA (M)
    M30b2a
    Y-DNA (M)
    R1a1-L657 (R-Y16494)
    mtDNA (P)
    M18a

    United States of America Canada
    Also, considering the variance within Indian groups, like how some of the Jewish or Shi'a groups will have a touch more discernible Caucasian/West-Asian ancestry, genetic recombination renders most of the variance we see, as unusual as it may be, within the realm of statistical noise. I've seen 5-10% swings between siblings. My own family has this random bit of "SW-Asian" which can hit 2-3% which is very unusual for north India but not if you consider it random variation possible from two parents with 1-1.5% SW-Asian. A lot of the random high levels of Mediterranean, Caucasian, SW-Asian or lower levels of something else on the individual/particular level could be chalked up to that.

    In my family's case, I averaged out like a dozen members and it's maybe 1% above average SW-Asian so it could be anything but that still doesn't rule out some recombination-related skewing that happened a few generations ago.

    Which is why I think we should reemphasize the importance of getting raw data which overlaps better with the calculators. The old platforms had 90+% overlap, the new ones are like 30-40% and there's a huge difference in results. If DNA.Land re-enables downloading of imputed data, I think everyone coming here who is tested on a newer genotyping chip should be advised to combine that raw data with imputed data before using it in calculators or even submitting it for G25.

    That and/or documenting the chips the samples are from and grouping them together in spreadsheets, etc
    Paternal - Y-DNA: J2b2* (J-M241) Z2432+ Z2433+ Y978+ (J2b2a2b1*) (Hidden Content ) (YFull: YF02959) (FTDNA Kit B6225), mtDNA: M18a* (FTDNA Kit 329180) (YFull: YF63773)
    Maternal- Y-DNA: R1a1a1b2a1a2c2d5a* L657+ Y7+ (R-Y16494*) (FTDNA Kit 311047) (YFull: YF68408), mtDNA: Hidden Content (FTDNA Kit B6225) (YFull: YF02959) (Mother's Mother's Father: R1a1a1b2a1a2c2* Y7+ Y29+ (R-Y29*) (FTDNA Kit 329181) (YFull: YF65256))

  6. The Following 2 Users Say Thank You to Dr_McNinja For This Useful Post:

     khanabadoshi (07-24-2020),  subzero85 (07-24-2020)

  7. #24
    Registered Users
    Posts
    2,622
    Sex
    Location
    New York
    Ethnicity
    South Asian
    Nationality
    American
    Y-DNA (P)
    J2b2-M241 (J-Y978*)
    mtDNA (M)
    M30b2a
    Y-DNA (M)
    R1a1-L657 (R-Y16494)
    mtDNA (P)
    M18a

    United States of America Canada
    Quote Originally Posted by pakistani View Post
    I am an Urdu speaker on both sides. To be more precise, and only including what I've been able to sufficiently verify:

    My mother's side was based in and around Delhi for centuries (since at least the 1300s). They primarily spoke Urdu and Persian and some knew Arabic as a scholarly profession. Ancestors not from the subcontinent came from Bukhara in the 1300s, Hadhramaut in the 1500s, and Baghdad in the 1700s. Endogamy was practiced on this branch quite a fair amount. I shared a fairly detailed family tree of my mother's side here if it may interest anyone or yield anything, but I'm fairly sure there weren't any Pashtuns on that side.

    My father's side is where I'm much less certain. There's the topic of discussion Mr. Hashmatullah Khan who was one of my great-great grandfathers, then there's another UP Khan I mentioned here who also claimed to have Pashtun ancestry. My direct line was a UP Syed, and the last great-great grandfather I know nothing about was named "Hakeem Faizzal Hassan." As far as I know everyone spoke Urdu on this side too.

    So that's what I'm working with. Seems like Urdu speakers in general are the mutts of the subcontinent.

    I actually did do DNA.land about two years ago, attaching the results.dna.land.png
    If/when DNA.Land re-enables downloading of data files, download your imputed VCF and use DNA Kit Studio to make a raw data file from that
    Paternal - Y-DNA: J2b2* (J-M241) Z2432+ Z2433+ Y978+ (J2b2a2b1*) (Hidden Content ) (YFull: YF02959) (FTDNA Kit B6225), mtDNA: M18a* (FTDNA Kit 329180) (YFull: YF63773)
    Maternal- Y-DNA: R1a1a1b2a1a2c2d5a* L657+ Y7+ (R-Y16494*) (FTDNA Kit 311047) (YFull: YF68408), mtDNA: Hidden Content (FTDNA Kit B6225) (YFull: YF02959) (Mother's Mother's Father: R1a1a1b2a1a2c2* Y7+ Y29+ (R-Y29*) (FTDNA Kit 329181) (YFull: YF65256))

  8. The Following 2 Users Say Thank You to Dr_McNinja For This Useful Post:

     pakistani (07-24-2020),  subzero85 (07-24-2020)

  9. #25
    Registered Users
    Posts
    468
    Sex
    Location
    London, UK
    Ethnicity
    Bihari, UPite and Bengali
    Nationality
    British
    Y-DNA (P)
    O2a2b1a1
    mtDNA (M)
    M6a1a
    Y-DNA (M)
    R1a1a-Z93, R-M560

    India Pakistan United Kingdom
    Quote Originally Posted by khanabadoshi View Post
    You need a plan of action to tackle this. Firstly, you to need establish or estimate timeframes. Do they match up with historical events that make that migration plausible? Secondly, you need to understand the relative relevance of a distant ancestor to your genetic profile and the possibility of its detection.



    THEORHETICALLY you need to test:
    Great-great Grandparent to detect an unadmixed ancestor 10 generations ago
    Great Grandparent to detect an unadmixed ancestor 9 generations ago
    Grandparent to detect an unadmixed ancestor 8 generations ago
    Parent to detect an unadmixed ancestor 7 generations ago
    Yourself to detect an unadmixed ancestor 6 generations ago.

    PRACTICALLY, you need to test:
    Great-great Grandparent to detect an unadmixed ancestor 8 generations ago
    Great Grandparent to detect an unadmixed ancestor 7 generations ago
    Grandparent to detect an unadmixed ancestor 6 generations ago
    Parent to detect an unadmixed ancestor 5 generations ago
    Yourself to detect an unadmixed ancestor 4 generations ago.

    The easiest thing to detect would be Iraqi ancestry from the 1700s -- it is the most recent in time and a stream of ancestry completely off the S/SC Asian genetic cline. Even though the Yemeni linkage is too far removed to detect on its own, the Iraqi addition later would compound the MENA stream of ancestry as a whole; this coupled with the fact that Muslims tend to practice cousin-marriages means that your MENA ancestry should be detectable at the very least in your grandparent(s) (if we are talking mid- to late- 18th century Baghdadi ancestor).

    Teasing out ancestry along the S/SC Asian genetic cline itself is tedious and challenging -- this is mostly because you are already on said cline, and not on the opposite end of the cline from a Pashtun, but rather somewhere in the middle of it. To put this into perspective and temper expectations... my grandmother is 75% Kho/Chitrali, my mother 37.5%, and myself 18.75% -- forget me... my MOTHER can be modeled without Kho OR any of the stand-ins (like Tajik or Pashtun). Let's say I didn't know what my grandmother was, I never tested her or my mother, and all I had was my own result... I'd never in a million years suspect Kho ancestry or assume it. I might have suspected a general stream of ancestry from the Northwest and I'd likely have attributed it to Pashtuns since they are the major population, but more likely than not, I'd just assume I was a Jatt or Gujjar (especially if I knew nothing specific of any of my grandparent's ancestry). Now the caveat is that this is relative... if your grandmother has the same profile as mine, but you are sitting in Uttar Pradesh... this stream of ancestry would make you more distinct from the population around you than it distinguishes me from the population around me in Punjab, and thus, it would be far more useful. You could come to much stronger conclusions than I could; in fact, you'd probably have to suspect a recent ancestor from further NW in South Asia, at the very least. However, even you would likely never assume Chitrali or even Burusho, or Shina... you might settle on Jatt or Kashmiri, or Tajik or Pashtun.

    So what to do? Distinguish between what you know and suspect. Assume all stories are true and sit there and test this. How? Well, you seem to have a strong family tree game going on... so assume so-and-so is 100% something and work your way down. Assume all non-related wives are local to wherever that ancestor is in time unless otherwise specified. Assign everyone something and work your way down to you. You'll have a theoretical sense of what you "should" be if "all the stories are true". Then run a bunch of admix calculators and see if the fits are decent with those populations alone. Eliminate, add, rework the theory... repeat.

    It is hard work. I am trying to figure out what my maternal Grandfather's grandmother was... and as of yet I can't. My best guess is she is from either Shamli in Uttar Pradesh, Bharuch in Gujarat, or Maharashtrian... and that is from 23andme relatives being related to specific people in my family only descended from this woman. Admix calculators have been relatively useless here, except that they point me eastward.
    Thank you so much for this Khanabadoshi, I know this is unrelated but I recently discovered that I have a 4th English cousin living in Cambridge, Massachusetts. Thanks to that chart I now know that we share between 1.5 to 3.5% of our DNA.

    More on-topic my 1st cousin in Delhi has a maternal ancestor of Afridi descent from Malihabad in Uttar Pradesh, India. Would it show up at all?.

  10. The Following User Says Thank You to deuterium_1 For This Useful Post:

     khanabadoshi (07-24-2020)

  11. #26
    Quote Originally Posted by khanabadoshi View Post

    So what to do? Distinguish between what you know and suspect. Assume all stories are true and sit there and test this. How? Well, you seem to have a strong family tree game going on... so assume so-and-so is 100% something and work your way down. Assume all non-related wives are local to wherever that ancestor is in time unless otherwise specified. Assign everyone something and work your way down to you. You'll have a theoretical sense of what you "should" be if "all the stories are true". Then run a bunch of admix calculators and see if the fits are decent with those populations alone. Eliminate, add, rework the theory... repeat.

    It is hard work. I am trying to figure out what my maternal Grandfather's grandmother was... and as of yet I can't. My best guess is she is from either Shamli in Uttar Pradesh, Bharuch in Gujarat, or Maharashtrian... and that is from 23andme relatives being related to specific people in my family only descended from this woman. Admix calculators have been relatively useless here, except that they point me eastward.

    This is exactly the use case that the "Ghost" creation feature in GenoPlot was designed for. You can start with the base assumptions derived from your ancestral paper trail and/or oral history and then use the resulting sample to verify or disprove specific potential ancestral populations by comparing against yourself.

    As an example



    Note that you can also use the "Subtract" feature to derive unknown ancestry from a given sample. You would simply subtract the
    known ancestries from the sample with unknown ancestry. This would leave you with a ghost that would represent the unknown ancestry. Plotting this ghost on a PCA, or running checkFits against suspected populations, will then identify the closest populations.
    Last edited by GenoPlot; 07-24-2020 at 03:39 PM.
    Explore your Genealogy

  12. The Following 11 Users Say Thank You to GenoPlot For This Useful Post:

     Ahmed Ali (07-24-2020),  Amber29 (07-24-2020),  deuterium_1 (07-24-2020),  discreetmaverick (07-24-2020),  Dr_McNinja (07-24-2020),  khanabadoshi (07-24-2020),  misanthropy (07-25-2020),  MonkeyDLuffy (07-24-2020),  Riverman (07-24-2020),  Sapporo (07-25-2020),  ThaYamamoto (07-24-2020)

  13. #27
    Registered Users
    Posts
    588
    Sex
    Ethnicity
    Deccani Muslim
    Y-DNA (P)
    E-V22
    mtDNA (M)
    M2B

    Quote Originally Posted by khanabadoshi View Post



    THEORHETICALLY you need to test:
    Great-great Grandparent to detect an unadmixed ancestor 10 generations ago
    Great Grandparent to detect an unadmixed ancestor 9 generations ago
    Grandparent to detect an unadmixed ancestor 8 generations ago
    Parent to detect an unadmixed ancestor 7 generations ago
    Yourself to detect an unadmixed ancestor 6 generations ago.

    PRACTICALLY, you need to test:
    Great-great Grandparent to detect an unadmixed ancestor 8 generations ago
    Great Grandparent to detect an unadmixed ancestor 7 generations ago
    Grandparent to detect an unadmixed ancestor 6 generations ago
    Parent to detect an unadmixed ancestor 5 generations ago
    Yourself to detect an unadmixed ancestor 4 generations ago.
    That's interesting. I thought I figured out my Yemeni ancestor was 4 generations ago, because I originally got 6.2% MENA on 23andme. Looking at your chart, he would be born roughly around 1860. I figured he was a Maratha mercenary due to the village he was from, but the Maratha army was disbanded and defeated by the British in 1818! That's wayy before he would theoretically be born if I believed he was 4 generations ago and nearly all my 6% MENA was from him. If he truly was one of those original mercenaries, then my MENA would have to decrease to 3% or less. I'm confused now lol. My other side has Syed lineages as well. I have to test a parent to get an idea of how much MENA is contributed from each side.

    Edit: to make things more confusing, my paternal cousin (same side as the Yemeni) who is half hispanic got 11% MENA on 23andme, I would guess 2-3% of that is from hispanic. My brother got nearly 9% MENA originally on 23andme.
    Last edited by misanthropy; 07-25-2020 at 08:10 PM.

  14. The Following 3 Users Say Thank You to misanthropy For This Useful Post:

     discreetmaverick (07-25-2020),  khanabadoshi (07-25-2020),  ThaYamamoto (07-25-2020)

  15. #28
    Registered Users
    Posts
    3,360
    Sex
    Location
    Chicago
    Ethnicity
    Baloch Kashmiri Uzbek Kho
    Nationality
    USA
    Y-DNA (P)
    R-Y17491 > R-YP4858
    mtDNA (M)
    A8a

    Pakistan United States of America
    Quote Originally Posted by misanthropy View Post
    That's interesting. I thought I figured out my Yemeni ancestor was 4 generations ago, because I originally got 6.2% MENA on 23andme. Looking at your chart, he would be born roughly around 1860. I figured he was a Maratha mercenary due to the village he was from, but the Maratha army was disbanded and defeated by the British in 1818! That's wayy before he would theoretically be born if I believed he was 4 generations ago and nearly all my 6% MENA was from him. If he truly was one of those original mercenaries, then my MENA would have to decrease to 3% or less. I'm confused now lol. My other side has Syed lineages as well. I have to test a parent to get an idea of how much MENA is contributed from each side.

    Edit: to make things more confusing, my paternal cousin (same side as the Yemeni) who is half hispanic got 11% MENA on 23andme, I would guess 2-3% of that is from hispanic. My brother got nearly 9% MENA originally on 23andme.
    This isn't some hard and fast chart. I just made it as a rough guide. I assumed a generation was 30 years apart. In reality, the year differences can vary alot between generation to generation. What if your grandfather is kid number 11 from the eldest born? 25 years between one generation and 50 years between the next, is possible. In our region, a generation could be 20-60 years apart on the male side and on the female side from as little as 15 years to as much as 45-50 years apart.
    Last edited by khanabadoshi; 07-25-2020 at 11:38 PM.
    “Chahar chez est tohfay Multan, Gard-o- Garma, Gada-o- Goristan”.

    Four things are the gift of Multan: Dusty winds, hot seasons, beggars and graveyards.




  16. The Following 3 Users Say Thank You to khanabadoshi For This Useful Post:

     misanthropy (07-26-2020),  Riverman (07-26-2020),  ThaYamamoto (07-25-2020)

Page 3 of 3 FirstFirst 123

Similar Threads

  1. Indians with pashtun ancestry?
    By Heteorchromia in forum Central
    Replies: 217
    Last Post: 11-22-2019, 03:26 AM
  2. Pashtun K13 calculator
    By Magnetic in forum Autosomal (auDNA)
    Replies: 21
    Last Post: 02-11-2019, 01:03 AM
  3. What is the difference between Afghan and Pashtun?
    By Alanson in forum General Sociology/Ethnology
    Replies: 16
    Last Post: 06-19-2018, 10:15 PM
  4. Replies: 152
    Last Post: 09-29-2016, 05:15 AM
  5. Replies: 56
    Last Post: 05-20-2016, 09:38 AM

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •