Forum:Methodology revolution of DNA and implications for this wikia

Forums: Index > Watercooler > Methodology revolution of DNA and implications for this wikia

It seems to me that the standards of extreme rigour necessary for teasing out the fact from the fiction in dearly held beliefs/ "family lore" about descendants would require this site to become less of a community site, and more a land where only the truly scholarly macho folks dare post anything.

I guess it rather depends on what you are interested in doing. Most folk doing genealogy are quite content just to get a lineage down that they think is theirs. I doubt anyone is intested in knowingly having a bad lineage, but most probably don't look too closely at the underpinnings of what they believe to be true. Its a very egalitarian hobby. Anyone can do, and unfortunately (perhaps) do. About 1% of the genealogists pursue things with a vigorous technical approach. They tend to honk off the other 99%. But I suspect the wiki is a mansion of many rooms, with a place for everyone, irrespective of approach. Bill 02:10, 4 September 2007 (UTC)

My observation is that the increasing cheap Y-DNA (male line) and mitochodrial mt-DNA (female line) tests revolutionize the standards for accepting data on parentage. Now, folks can post what they heard from grandmothers, etc. and post it as such. Later, ancestors can come along, do the test and if they get a positive match, then much more credence can be give to what would previously been regarded as unsubstantiated myth.

Anyway, it is my intention to document where I am getting information but not get hysterical about securing the primary source material just yet. If I can verify via DNA testing some of these connections, I can go deeper into the family tree a lot lot faster.

-Mak 23:58, 2 September 2007 (UTC)

Well said, Mak! Even with DNA work, there are drawbacks, including substantial room for error. What Granny said may contain at least a quarter-truth and can lead us in the right direction. Robin Patterson 12:09, 3 September 2007 (UTC)

The drawbacks are that it is only good for Verifying pure male (son to father to grandfather but not son to mother to grandfather) or with mDNA, pure female lines. As for accuracy, a 25 marker match with a relative is as strong as the strongest case made with the most stringent standards for documentation. At the end of the day, it is as the proverb goes: Maternity is a matter of fact. Paternity is a matter of opinion. Well, maybe not anymore. As for non paternal events, well- most of these folks are dead and buried and there are no feelings to hurt, so ok some of the fathers may have not been the real fathers. Big deal.

Well, perhaps not a big deal to you, but I know folks who are intensely troubled by some of the revelations that have occurred in their family history for ancestors dating back into the early 1800's, and in one case into the early 1700's. It may not bother you, and it does not bother me, but it does bother some folks. Bill 02:04, 4 September 2007 (UTC)

Point taken. Many are interested in genealogy because they are sort of into ancestor worship in a way- that is they feel that it is important to amplify the respect and rememberance of the great things that our ancestors did. We want to assume the best of our ancestors, and there is a temptation to avoid any cognitive dissonance that conflicts with our grand pictures of them. In my opinion, being realistic about what they did brings us closer to their humanity, keeps them alive in a way. The good and the bad. You turn them into idols, and they become like stone- They are made immovable with the fantasies we ensnare them in- they can no longer speak to us. They are dead. -Mak 09:51, 4 September 2007 (UTC)

Yes, for some, there is a certain amount of ancestor worship. Indeed, in some cultures I can think of genealogy is an important adjunct to the formal religious practice. The Mormon's are an obvious example, but I'm really thinking of other cultures where ancestor worship IS the religion. And I've had conversations with one genealogist (whose contributions to our family studies I highly value) for whom the received knowledge of the family was all important. Indeed, this individual made it quite clear that if we DID learn something that contradicted the received knowledge, he was not interested in hearing about it. It was more important for him that his understanding of his family connections remain intact, than that those connections be accurate. Its a view I do not share, but it is his view, his choice, and I respect his preferences. Bill 12:59, 4 September 2007 (UTC)

It will be a big deal for motivated researchers though. One gal on NPR's science friday was saying that their data so far using these markers is that non paternal events are detected for 15% of the cases. Seems small, but statistically it adds up when you are going back 8 or 9 generations. What happens to my interest in Messerli family ancestors in 17th century Switzerland when I find out that my great great great grandmother was not baking pies while grandpa was in the fields?

One thing that is not clear to me is if it is possible to quantify the degree of uncertainty created over several generations when very large marker sets that are currently available (eg 65). If you could, it would be nice so that folks could say- "well, we have only a few recollections recorded for this parentage, but the dna match indicates a probability of 85%, so that's good enough. -Mak 20:04, 3 September 2007 (UTC)

I haven't seen the statistical projections for the 65 marker tests, similar to those readily available for the 12, 25 and 37 marker test. I would guess, though, that the improvement it provides is a little less than double that obtained for the 37 marker test, based on earlier projections of what it would take to get a commensurate improvement in confidence. However, are you really happy with being wrong 15 times out of a hundred? Bill 02:04, 4 September 2007 (UTC)

What I meant was that matching is really not a boolean condition. The more markers that don't match, the lower the probability of a recent common ancestor, but it's only a probability. With a complete match, there is extremely high confidence in a close common ancestor. You can take that to the bank- no doubts about non paternal events there, that no amount of paperwork can ever disprove in a million years.

Are traditional genealogical methods obsolete? Of course not. Because though you may have established that you have close common ancestors, you may be related by means of a common father or grandfather. Only honest rigorous research can take you to the finish line.

There is a paper of interest on the DNA:Riggs page. Contrary to intuitition, the larger number of markers doesn't do much for the accuracy between two given individuals. It does have a use though, and it would be good for folks to get large sets if they want to research deep pedigree trees. A larger number of markers gives a better footprint from which to establish a common denominator signature determined from known ancestors established through traditional genealogical means.-Mak 09:32, 4 September 2007 (UTC)

The issue with increaseing the number of markers is related to cost-benefit. It roughly takes a doubling of the number of markers to get a commensurate increase in confidence. YOu get substantial improvement in adding an additional 13 markers when you go from 12 to 25, but you don't get quite the same level of improvement if you add another 12 markers going to 37. To get the same level of improvement you need to add an additional 50 markers. The net effect is that commensurate levels of improvement are progressively more and more expensive. At an early stage in the evolution of this FTDNA was recommending 37 markers because it got you the most bang for the bucks. They added the 62 marker test, I suspect, simply because people wanted more. Haven't been many takers on that, probably because of the cost.

I would guess that the cost will drop as demand for DNA testing increases---do to more factors than genealogy. At some point a DNA test at the Doctor's office is going to be a routine part of a check up. When that happens, I think we can expect that the volume will drive down the test cost t the point where $100 will get you far more than a simple YDNA test of 67 markers. It will be several years more before this gets to be routine, but it will happen. Then we will have a field day. Among other things, the specialized haplogroup and haplotype testing will become trivial (I hope). I'm not much interested in the deep genealogy that's embodied in the haplogroup results, but as it turns out that kind of data (specifically from the actual STR types of tests) can real condition and inform YDNA matches---I know of one case at I think 37 markers, where a one-off rusult with the same surname, could be rejected because haplotypes mismatched. (Haplotypes tests are based on different suites of markers than the YDNA test itself; the folks at FTDNA can usually guess as to what haplogroup/type is involved based on the 12 marker test results, but its a guess. In this case, I was fortunate that there were actual STR haplogroup test results.) Bill 12:59, 4 September 2007 (UTC)