Size of DNA data bases

[Guy Etchells]

snip

We do indeed inherit exactly 50% of our DNA from each parent, because we inherit one of each chromosome pair from each. And that inevitably means that we inherit half of each of their DNA.

The only exception to this is the 23rd chromosome, which determines gender. The X chromosome, which both genders have is larger than the Y chromosome which only men have, so a man inherits a longer length of DNA in the 23rd chromosome from his mother than his father. But that fact isn't significant to what we are discussing here.

Eric you seem to know more than Rolf Kohl, Dr.rer.nat. (Ph.D) Ecology & Biology, Ludwig Maximilian University of Munich, for instance,  who states-
“child inherits EXACTLY 50% of the chromosomes from each of both parents. All mitochondrial genes are inherited only from the mother.
Thus a child receives only close to half of the genes from the father and a little bit more than half of the genes from the mother.”

We inherit approximately (but not exactly) 25% from each of our 4 grandparents (i.e. 2 generations back), approx 12.5% from 3 generations back, approx 6.25% from 4 generations back, approx 3.12% from 5 generations back, etc. The further back we go, the smaller the percentage and the greater the potential for variation.

Which is what I was saying due to DNA dropout (this is where the DNA of a particular ancestor is not passed down to a child, that particular DNA may be passed on to a sibling but the child’s descendants will never carry the “missing” DNA.

I can only conclude that you have taken an unnecessarily sceptical and not fully accurate view of DNA testing. I hope you find the links I have given helpful in building your understanding, as they have helped me. Thanks.

Far from “taking an unnecessarily sceptical and not fully accurate view of DNA testing” I have carried out in depth study of many hundreds of scientific evaluations of DNA over the last 17 or so years. I approached the subject hoping it could prove to be a very useful genealogical tool, but my conclusions are the current offerings contain more hype and miss information than useful data.
The Ancestry offerings are at present going to produce hundreds if not thousands more inaccurate Ancestry trees due to the many false positives that the naive are going to cling onto to “prove” their inaccurate trees.
As I have said in the past when the science develops it may prove to be of use in genealogy but at present it is far too inaccurate and too expensive to be of any use to the bulk of genealogists.

That is not to say it cannot be used as an additional tool by careful researchers who have a well researched lineage but for the customer base the adverts are targeting it is as useful as a pin and a telephone directory
Cheers
Guy

[Eric Hatfield]

That is so cool, you are a mine of good information. Thanks.

I actually visited the DNA geek blog just a couple of days ago and asked about the different companies' coverage in different countries, so perhaps this survey is partly as a result.

I hope she gets enough participants to get some reasonable results.

[Eric Hatfield]

Hi Guy, You say "Eric you seem to know more than Rolf Kohl, Dr.rer.nat. (Ph.D) Ecology & Biology, Ludwig Maximilian University of Munich, for instance,  who states-", but he said the same as I did, except I didn't mention mtDNA, which is quite separate from the 23 chromosomes, but which men inherit just as much as women do. Your quote is about genes, which don't account for all DNA, but only a small part of it.

Which is what I was saying due to DNA dropout (this is where the DNA of a particular ancestor is not passed down to a child, that particular DNA may be passed on to a sibling but the child’s descendants will never carry the “missing” DNA.

Yes, there is a drop off. Nevertheless, some DNA is generally passed down from ancestors in the first 5 generations back. My understanding is that for a person to receive no DNA from such a close ancestor would be rare. So there is almost always enough DNA to form a match between two people with a common ancestor in the last five generations.

As I have said in the past when the science develops it may prove to be of use in genealogy but at present it is far too inaccurate and too expensive to be of any use to the bulk of genealogists.

Ah well, I guess you can ignore it. Meanwhile, so many other people, including a few on this thread, have found it useful, actually indispensable. But let's not go over that ground again shall we?

[hurworth]

Yes, there is a drop off. Nevertheless, some DNA is generally passed down from ancestors in the first 5 generations back. My understanding is that for a person to receive no DNA from such a close ancestor would be rare. So there is almost always enough DNA to form a match between two people with a common ancestor in the last five generations.

Not quite true.  Yes, it would be rare for a person not to inherit any DNA from an ancestor five generations back (a gtgtgt-grandparent).  But just because you inherit DNA from an ancestor doesn't mean your fourth cousin who descends from the same couple will have inherited the same DNA from them that you have.

This table has the likelihood of matching different cousins
https://isogg.org/wiki/Cousin_statistics

[sugarfizzle]

Guy, you say "That is not to say it cannot be used as an additional tool by careful researchers who have a well researched lineage but for the customer base the adverts are targeting it is as useful as a pin and a telephone directory"

I think we are in agreement there. I use it as an additional tool and have found it very helpful. My lineage is carefully researched and has not a single reference to anyone else's tree.

Those targeted for and promised their ethnicity will be very disappointed, those who have a genuine interest in genealogy and who are willing to research will most probably not be disappointed.

Inbetweeners, who have a vague interest in their ancestry but are not willing to do any research into either their tree or their DNA matches, will also probably be disappointed.

For people like myself, prepared to spend several hours a day in researching both my tree and my matches, DNA has become an invaluable tool.

Regards Margaret

[Eric Hatfield]

Not quite true.  Yes, it would be rare for a person not to inherit any DNA from an ancestor five generations back (a gtgtgt-grandparent).  But just because you inherit DNA from an ancestor doesn't mean your fourth cousin who descends from the same couple will have inherited the same DNA from them that you have.

Hi Hurworth, yes you are right. My first statement was  correct, but only a percentage of 4th cousins will match. I mentioned some of the percentages before, and your reference provides them too. Thanks for the correction.

My point was that while we miss some matches, we also get plenty, which is better than nothing. (It's a bit like paper records. Some church records for example have been lost, some are still available, but that doesn't stop us using what we can find.)

[Eric Hatfield]

Removed because I got it wrong. Sorry.

[Pheno]

It seems quite clear to me really.

There's nothing wrong with a dna test for ethnicity for anybody who thinks they would like to have some idea of their origins.  Even if these are not accurate it still satisfies a need in them and they probably won't ever return to check it out - hence the number of Ancestry tests without attached trees.

For those that are interested and take the test - ethnicity predictions are a bit of  bonus.  However, matching with others with similar dna is what its all about - but for the true genealogist it won't be taken as gospel without proof of a paper trail.  Dna results simply provide a possible clue about ancestors and then you need to prove it.  Thats the thrill of the chase isn't it and don't verify a dna connection without the paperwork to prove it.

Pheno