DNA Why I urge caution

[Sandra H]

I am still very new to researching my family history but I can see how DNA matches could possibly help provide some useful information to confirming some family lines.

[Lubana]

People who have been able to compare their DNA to other close family members on a site like Ancestry know just how accurate that is.  Parents share about 50% of their DNA with their children and, likewise, full siblings share roughly 50% as well.  That comes out to about 3, 500 centimorgans, give or take.  Almost needless to say, the more distant the relationship, the fewer cMs.  This is a good checking source, if in doubt.

https://dnapainter.com/tools/sharedcmv4/1089

Worrying about whether an identical twin brother of someone might have been your father is rather pointless, IMO, as it is so unlikely to have been the case,  It is only identical twins that have the same DNA.  Fraternal twins are like any other full siblings when it comes to DNA.  50% of it will be different. 

The gray area involves ethnicity.  While one company will identify you as 30% British, say, another will assign you perhaps 38%.  All you can deduce from that is that you have a significant amount of British ancestry.  Another company--not Ancestry--at one time assigned me some small bits of interesting ethnicity--and ended up taking that away for a reason I have yet to understand.  But, if one can acknowledge a lack of complete accuracy in that aspect of ones DNA results, one can live with it.  It is not to be confused with the accuracy of relationships between people, that's all.  The reliable testing companies look at hundreds of thousands of markers, not just a few.  You can bank on the results.

[davidft]

You can bank on the results.

That's a "brave" thing to assert. Not even the genealogical DNA companies themselves claim that.

[Guy Etchells]

People who have been able to compare their DNA to other close family members on a site like Ancestry know just how accurate that is.  Parents share about 50% of their DNA with their children and, likewise, full siblings share roughly 50% as well.  That comes out to about 3, 500 centimorgans, give or take.  Almost needless to say, the more distant the relationship, the fewer cMs.  This is a good checking source, if in doubt.

https://dnapainter.com/tools/sharedcmv4/1089

Worrying about whether an identical twin brother of someone might have been your father is rather pointless, IMO, as it is so unlikely to have been the case,  It is only identical twins that have the same DNA.  Fraternal twins are like any other full siblings when it comes to DNA.  50% of it will be different.

It may be pointless to you but I can assure you that when a mother just before she dies of cancer tells one of her daughter’s her father is really her grandfather that daughter does not think it pointless. In the above case the mother and her three elder daughters were living in the grandfather’s home while her husband was serving a prison sentence. There is very little chance the mother did not tell the truth, unless the daughter was born very premature or late but there was no hint of that.
Every DNA expert I have spoken to agrees it would be very difficult to discover the birth father via DNA.

The gray area involves ethnicity.  While one company will identify you as 30% British, say, another will assign you perhaps 38%.  All you can deduce from that is that you have a significant amount of British ancestry.  Another company--not Ancestry--at one time assigned me some small bits of interesting ethnicity--and ended up taking that away for a reason I have yet to understand.  But, if one can acknowledge a lack of complete accuracy in that aspect of ones DNA results, one can live with it.  It is not to be confused with the accuracy of relationships between people, that's all.  The reliable testing companies look at hundreds of thousands of markers, not just a few.  You can bank on the results.

I think my views on ethnicity results are well known.

With regards to the accuracy of DNA

The population of London in 2018 was 8,787,892 the largest worldwide DNA database is estimated at 12,777,778 in other words half as much again as 1 city or if you compare this to the world population of 7.6 billion (in 2018) in other words a small proportion.

A lot of claims are made about DNA but at present it is still very much a tool in the genealogists toolbox that relies heavily on other tools to produce results.

Cheers
Guy

PS All the daughters (5) have had their DNA tested, the older generations are all deceased, mother, father & grandparents.

[Lubana]

You can bank on the results.

That's a "brave" thing to assert. Not even the genealogical DNA companies themselves claim that.

Actually, they do.  Even paternity tests, involving fewer markers, claim close to 100% accuracy.  The thing is, if brothers may be involved, one ought to notify the tester.  Trust me, if they test enough markers, they will distinguish between two brothers or their father.  Have you ever seen a DNA testing company advertise "We may not be all that accurate"?

[Lubana]

People who have been able to compare their DNA to other close family members on a site like Ancestry know just how accurate that is.  Parents share about 50% of their DNA with their children and, likewise, full siblings share roughly 50% as well.  That comes out to about 3, 500 centimorgans, give or take.  Almost needless to say, the more distant the relationship, the fewer cMs.  This is a good checking source, if in doubt.

https://dnapainter.com/tools/sharedcmv4/1089

Worrying about whether an identical twin brother of someone might have been your father is rather pointless, IMO, as it is so unlikely to have been the case,  It is only identical twins that have the same DNA.  Fraternal twins are like any other full siblings when it comes to DNA.  50% of it will be different.

"It may be pointless to you but I can assure you that when a mother just before she dies of cancer tells one of her daughter’s her father is really her grandfather that daughter does not think it pointless. "

Hold on--what does this have to do with what I said about identical twins?  In fact, I fail to see your point at all. 


"In the above case the mother and her three elder daughters were living in the grandfather’s home while her husband was serving a prison sentence. There is very little chance the mother did not tell the truth, unless the daughter was born very premature or late but there was no hint of that.
Every DNA expert I have spoken to agrees it would be very difficult to discover the birth father via DNA."

I have no idea what experts you have spoken to--but difficult doesn't mean impossible.  A man's autosomal DNA is not a clone of that of his father.  In autosomal DNA, everyone has two alleles at every marker or locus, one donated by the father and one by the mother.  Obviously a father and his son do not have the same mothers.  Their DNA will look different, even though the son received an allele from the father at every marker.  What will look the same is their y-DNA, which is a different story.  Let's say Dad has these alleles, shown as numbers at a marker----12/23.  His female partner has 8/10.  Son will receive a number from each parent out of a possibility of four combinations.  Let's say he receives  10/23.  That illustrates why his DNA is not the same as that of his father.  Now Son can pass down nothing but 10/23 at that marker to his own children.  But Dad can only pass down 12/23 no matter what.  It doesn't take a rocket scientist to figure out that if Offspring has a 10 at the same marker, combined with another number from his or her Mom, it will not have come from Dad--but Son.

The gray area involves ethnicity.  While one company will identify you as 30% British, say, another will assign you perhaps 38%.  All you can deduce from that is that you have a significant amount of British ancestry.  Another company--not Ancestry--at one time assigned me some small bits of interesting ethnicity--and ended up taking that away for a reason I have yet to understand.  But, if one can acknowledge a lack of complete accuracy in that aspect of ones DNA results, one can live with it.  It is not to be confused with the accuracy of relationships between people, that's all.  The reliable testing companies look at hundreds of thousands of markers, not just a few.  You can bank on the results.

I think my views on ethnicity results are well known.

With regards to the accuracy of DNA

The population of London in 2018 was 8,787,892 the largest worldwide DNA database is estimated at 12,777,778 in other words half as much again as 1 city or if you compare this to the world population of 7.6 billion (in 2018) in other words a small proportion.

A lot of claims are made about DNA but at present it is still very much a tool in the genealogists toolbox that relies heavily on other tools to produce results.

Your views on DNA may be well known [but not to me] so forgive me if I question *their* acciracy.  The population of London now or at any other time has nothing to do with the accuracy of DNA testing.

[davidft]

Actually, they do.

OK prove it. As I said not even the companies themselves make the claim you did.

Have you ever seen a DNA testing company advertise "We may not be all that accurate"?

Now that is just being silly in my opinion. That said I think its a good time for me to leave this conversation as I have a feeling no good will come of it.

[Guy Etchells]

The gray
I have no idea what experts you have spoken to--but difficult doesn't mean impossible.  A man's autosomal DNA is not a clone of that of his father.  In autosomal DNA, everyone has two alleles at every marker or locus, one donated by the father and one by the mother.  Obviously a father and his son do not have the same mothers.  Their DNA will look different, even though the son received an allele from the father at every marker.  What will look the same is their y-DNA, which is a different story.  Let's say Dad has these alleles, shown as numbers at a marker----12/23.  His female partner has 8/10.  Son will receive a number from each parent out of a possibility of four combinations.  Let's say he receives  10/23.  That illustrates why his DNA is not the same as that of his father.  Now Son can pass down nothing but 10/23 at that marker to his own children.  But Dad can only pass down 12/23 no matter what.  It doesn't take a rocket scientist to figure out that if Offspring has a 10 at the same marker, combined with another number from his or her Mom, it will not have come from Dad--but Son.

It would certainly make things easier if it was possible to test the husband and his father, but both of these people had died before DNA testing became generally available as was the mother but perhaps you missed that small point.

The gray area involves ethnicity.  While one company will identify you as 30% British, say, another will assign you perhaps 38%.  All you can deduce from that is that you have a significant amount of British ancestry.  Another company--not Ancestry--at one time assigned me some small bits of interesting ethnicity--and ended up taking that away for a reason I have yet to understand.  But, if one can acknowledge a lack of complete accuracy in that aspect of ones DNA results, one can live with it.  It is not to be confused with the accuracy of relationships between people, that's all.  The reliable testing companies look at hundreds of thousands of markers, not just a few.  You can bank on the results.

So what do you imagine British ethnicity is, Norman, Viking, Gaelic, Germanic or one more of the other ethnic groups that have interbred with those living on the British Isles over the last couple of thousand years. There is no such thing as British ethnicity. But as you mention if you ignore the fact that the tests are not 100% (Which even the DNA companies admit) accurate you can bank on the results.

Your views on DNA may be well known [but not to me] so forgive me if I question *their* acciracy.  The population of London now or at any other time has nothing to do with the accuracy of DNA testing.

The reason I mention population is at present DNA databases have only just scraped the surface of the population of the world. You may assume everything is determined and indisputable but even the DNA companies themselves realise they are still learning and still developing theories.
An example of this is shown by their changing ethnicity estimates.
Cheers
Guy

[Lubana]

The gray
I have no idea what experts you have spoken to--but difficult doesn't mean impossible.  A man's autosomal DNA is not a clone of that of his father.  In autosomal DNA, everyone has two alleles at every marker or locus, one donated by the father and one by the mother.  Obviously a father and his son do not have the same mothers.  Their DNA will look different, even though the son received an allele from the father at every marker.  What will look the same is their y-DNA, which is a different story.  Let's say Dad has these alleles, shown as numbers at a marker----12/23.  His female partner has 8/10.  Son will receive a number from each parent out of a possibility of four combinations.  Let's say he receives  10/23.  That illustrates why his DNA is not the same as that of his father.  Now Son can pass down nothing but 10/23 at that marker to his own children.  But Dad can only pass down 12/23 no matter what.  It doesn't take a rocket scientist to figure out that if Offspring has a 10 at the same marker, combined with another number from his or her Mom, it will not have come from Dad--but Son.

"It would certainly make things easier if it was possible to test the husband and his father, but both of these people had died before DNA testing became generally available as was the mother but perhaps you missed that small point."

You never mentioned that all were deceased, so what was there to miss?  And how is all that the fault of some testing company?  Wasn't that your original point--their so-called lack of accuracy?  If the possible fathers are deceased, then there can't be a paternity test.  There is the possibility of autosomal testing and receiving some DNA relatives, the distance of relativity being calculated.  Let's say the deceased possible father, old Dad, had some other children.  What relationship of Offspring to them?  Or old Dad had some siblings with their own offspring--same question.  In that case, none of the potential fathers need to be alive.  All that's required is luck.

[snip of irrelevant stuff about ethnicity from Guy Etchells]