Author Topic: DNA 50:50 ?  (Read 277 times)

Offline JMB1943

  • RootsChat Member
  • ***
  • Posts: 147
  • Census information Crown Copyright, from www.nationalarchives.gov.uk
    • View Profile
DNA 50:50 ?
« on: Sunday 05 February 23 16:25 GMT (UK) »
My wife has All-Irish ancestry, except a great-grandmother from Switzerland.

As expected for a 50:50 DNA contribution, my wife’s DNA shows 12% North European (100—>50—>25—>12).

However, if the DNA is not a 50:50 contribution at each generation, such a 12% N European match is theoretically possible by episodes of excessive and minimal contribution from the Swiss lady’s genes.

For instance, 100—>80—>40—>12 which has only a single 50:50 contribution.

Are such wide swings from 50:50 known or common?

Regards,

JMB

Ballard/Hew/Mitchell/Mountcastle/Palmer/Spicer/White

Offline phil57

  • RootsChat Veteran
  • *****
  • Posts: 644
  • Census information Crown Copyright, from www.nationalarchives.gov.uk
    • View Profile
Re: DNA 50:50 ?
« Reply #1 on: Sunday 05 February 23 18:22 GMT (UK) »
Your receive 50% of your DNA from your father and 50 % from your mother, although in reality the ratios are not exactly equal, but near enough. They in turn received 50% of their DNA from each of their parents, but it does not follow that you received equal amounts of each grandparents' DNA. The DNA that you inherit from each of your parents is an entirely random combination of the DNA that they inherited from their parents. Consequently the amount of each grandparents' DNA that you inherit is not divided into four equal shares. You could have 40% or greater of your paternal grandfather's DNA, and 10% or less of your paternal grandmother's for instance, with a similar inequality between the inherited amounts on the maternal side.

To complicate matters further, you are referring to ethnicity ESTIMATES rather than direct familial inheritance. Ethnicity is an inexact science, being generous with the description, and the amount given by Ancestry is an estimation. So if you click on the headline figure of 12% given for your wife's North European ethnicity ESTIMATE (that word again) you will be shown a more detailed range of probability, which might explain that the estimate of 12% comes from a probability range of 0% to 23% (example figures plucked out of thin air for this example, but 0% is not unusual for fairly low estimates). In other words, your wife appears to have regions in common with Ancestry's North European reference population of up to 23% of her DNA, but Ancestry cannot be completely certain about that, and it might also be that she has absolutely no regions in common with that reference population at all, but as a result of that probability range they have come up with a figure of 12%.

So yes, such wide swings are not only common, but entirely expected, and how much credibility you give to ethnicity estimates is entirely your decision. But they are inexact by the very nature of being estimates derived from a probability range (and best not to be taken too seriously, particularly when you get into the exactness or otherwise of percentages quoted).

Additionally, ethnicity estimation is based on patterns believed to be common to general populations, not just immediate ancestors of a specific individual, so the estimated percentage/range given refers to all possibly matching patterns in the tested individuals' DNA and is not necessarily restricted to inheritance from a single parent/grandparent/GGP etc. It could just as easily be an overall estimation from combined inheritance on two or more ancestral lines. Granted, Ancestry's Sideview technology attempts to split it between paternal and maternal lines, but firstly that is a computer algorithm - again not an exact science - and it can't indicate whether a particular ethnicity shown on only the maternal or paternal side stems from a single grandparent, or both.

Ethnicity estimates are a completely different matter to segment matching between directly related individuals who descend from a common ancestor, which is a much more exact and provable science, and how we can identify people who are related to us.
Stokes - London and Essex
Hodges - Somerset
Murden - Notts
Humphries/Humphreys from Montgomeryshire

Offline JMB1943

  • RootsChat Member
  • ***
  • Posts: 147
  • Census information Crown Copyright, from www.nationalarchives.gov.uk
    • View Profile
Re: DNA 50:50 ?
« Reply #2 on: Sunday 05 February 23 20:37 GMT (UK) »
Phil,

Thank you for that very lucid explanation of DNA inheritance and the ethnicity that can be guesstimated from it.
I shall learn to tread very lightly in that particular jungle!

Regards,

JMB
Ballard/Hew/Mitchell/Mountcastle/Palmer/Spicer/White

Offline Biggles50

  • RootsChat Veteran
  • *****
  • Posts: 943
    • View Profile
Re: DNA 50:50 ?
« Reply #3 on: Sunday 05 February 23 22:01 GMT (UK) »
This thread should be Pinned for Phil’s detailed explanation which can then easily be referenced whenever there is another Ethnicity query.

We have so many posts by people who seem hung up on Ethnicity when it is an Estimate, I would further describe it as a Guestimate by Ancestry.