New Haplogroups for Old?

[john franks]

Looking at the ancient migration routes and the present day distribution of haplogroups around the world it seems odd to me that at the end-points of the migrations certain haplogroups dominate, for example R and derivatives in the British Isles and M in New Zealand. In both of these areas the precursors to R and M respectively may be present but in relatively small numbers. It just appears to me that whenever a mutation occurs, that in the modern day is recognized as a new haplogroup, then the odds should be against the man bearing the new marker multiplying at the expense of others who have the original dna. I can understand how the new marker can be used to track the movement of peoples around the globe but not how it comes to dominate.

I may have lost the plot somewhere but I would be grateful for comments

[Deirwha]

hmmmm

I think your comment is the proverbial tip of an iceberg ... that much goes into it which you understand but I am not sure if I understand. Not a criticism but a reflection.

I think if I do understand from where you are coming we may have a thought in common in mind. I have not understood how we can project the human journeys, stretching back over allegedly 60,000-200,000 years based almost entirely on the small segment of today's population who are willing to pay for tests. That does not seem a valid sample much less a controlled experiment. I do understand for a "control" one would need to test en masse the remains of the departed in strata and 1) techniques and DNA have not achieved the capability of doing that, and 2) the process is expensive and very few funding sources are out there to fund it. However, it has been a source of puzzlement and anxiety for me as I invest my money into these tests.

Why do I invest the money? Beyond my natural curiosity I would have to say it is the credibility National Geographic has with me. Because they are behind it I am trusting there is a process here that yields useful results even though I do not understand how that could be. I find the folks at FTDNA very pleasant and have met folks on this Forum and outside of it who strike me as credible people and certainly not cons. But it really is the reputation of National Geographic that has persuaded me to put my interrelated skepticism and confusion into abeyance and keep on with the process.

[cacio]

john:

I think the point is that the mutations we're considering are just a small subsets of all existing mutations in the Y. These SNPs were chosen exactly because they were common and easily recognized. What you call the "precursors" also have some mutations somewhere - but simply, their mutations haven't been looked into.

Start from the initial population. Everybody will have a number of random mutations in different parts of the Y. One of these people then leaves and goes to inhabit a new land, where his offsprings end up as the majority of the population. Because of the way SNPs are found and described, scientist will then pick up the SNP characterizing this individual. All the mutations that have occurred in the other people are not picked up and described.

Deirwha:
usually, when studying migrations and the like, scientific papers use selected and carefully chosen samples of people. These estimates are not based on ftdna customers.

cacio

[Clochaire]

Looking at the ancient migration routes and the present day distribution of haplogroups around the world it seems odd to me that at the end-points of the migrations certain haplogroups dominate, for example R and derivatives in the British Isles and M in New Zealand. In both of these areas the precursors to R and M respectively may be present but in relatively small numbers. It just appears to me that whenever a mutation occurs, that in the modern day is recognized as a new haplogroup, then the odds should be against the man bearing the new marker multiplying at the expense of others who have the original dna. I can understand how the new marker can be used to track the movement of peoples around the globe but not how it comes to dominate.

I may have lost the plot somewhere but I would be grateful for comments

Well, replacement of the "old" haplogroup by the "new" haplogroup could represent the result of lots of individually independent trends. Not being a demographer, I could only speculate at very rough means of estimating the significance of any individual trend, but here's one thing to think about:

Maybe it is the case that there were long bottle-necks in growth of human populations and migration. Maybe climate changes favorable to population growth at the edges of human settlement that you describe, like Europe and New Zealand, coincided with the occurrence of origination of "new" haplogroups?

Here's an over-simplified narrative, but hopefully it illustrates my principle thoughts clearly:

For millenia, our ancestors lived in Africa, with a comparatively stable and simple technology level that allowed only a very fixed level of population.

As time goes on, 2 phenomena occur independently, but within a proximate time frame: 1.) Climate changes favorable to migration out of Africa; 2.) Ocurrence of SNP mutations creating "new" haplogroups.

Maybe then, as now, only a relatively small percentage of the population felt enough incentive to migrate out of their homeland. Maybe the descendants of the migrants experience what I think is called a 'founder effect'. Because the numbers of the original migrants were small, genetic diversity was lower than in the homeland--thus more "predominant" in their new homes.

The divergence is made to appear even more striking because, lacking the same high level of economic limit on reproductivity upon migration to previously unpopulated Europe and Asia, within a very short period of time, the descedants of the "new" haplogroups become as numerous or more so than the descedants of the "old" haplogroups who remained in the homeland.

So maybe it is not a matter of "new" haplogroups eliminating or replacing "old" haplogroups, but finding opportunities for faster-growth in previously unsettled regions.

Jack

[Deirwha]

Caico

Thank you for that point. It is good to know that. They are, on the other hand, still people carefully chosen from people living today, I think. Or maybe I did not understand what I was reading. That is possible. Then there is the conclusion in one well known study that S29 represents 1.8% of the population of Russia based on one man with a German name. Your point that these are not self-selected subjects but screened and chosen subjects is a very good one. That does increase their credibility. However, I still hesitant to embrace wholeheartedly conclusions about yesterday based on subjects living today studied in small samples. Admittedly, big communication gap here because I suspect everything you would need to tell me about methodology for me to understand would be very technical and over my head. I remain "in this" because of the credibility of National Geographic, not because the explanations have yet made sense to me. As I say, the fault is undoubtedly my non technical mind and not the correctness of the explanations.

[cacio]

I think your point on small samples is right. I was just referring to what you said about "based almost entirely on the small segment of today's population who are willing to pay for tests.", ie willing to pay for test part. People who are willing to pay could be unrepresentative of the whole population, for many reasons. One example about Italy for instance: the majority of Italian Americans are from Southern Italy, and right now there are more Italian Americans who have purchased tests than Italians living in Italy. If you look at paying FTDNA customers, the percentages for those with Italian ancestry will be skewed towards S Italy.

However, what I meant is that scientists try to sample different people from different populations, so this bias is less relevant. But you are absolutely right about sample sizes. Even the best papers do not have more than a few hundred observations. That could be enough to determine the main haplogroups, but for haplogroups that have smaller percentages in the area, the estimates will have large confidence intervals. As you say, if you have a sample of 100 people, then 1% is just one person. If by chance you don't get anybody, you get 0%. Viceversa, there may be just a handful of people among millions in a certain haplogroup, and yet, if by sheer chance one of them is picked, you get 1%.

cacio

[Deirwha]

sample size

Yes, the same dynamic you mention about Italy is true of England, also.

My obsession with the absence of data on remains goes back to my pre genealogical interests, which included physical anthropology, archaeology and the history of particularly the ancient near east. I have had the need for controlled experimentation even in these fields drummed into my head for years. Also, there are some, oh, discrepancies is too strong a word, some theories of migration and cultural patterns based on linguistics and the archaeological record that need to be tied off. In general, our dates for types are either too young or too old depending on the association. I have wondered if they are not to young. However the newer studies seem to be suggesting they are way too old (ie, R1b at 3500 years not 18,500 while I was thinking 30-40,000). Whichever is the case, either 3500 or 30-40,000 seem to me to match up better than 18,500, BUT how will we be able to look ourselves in the mirrors and claim we know without stratified sampling of remains?

That is really my final point, pruned of the details. I just don't get how we can derive conclusions about ancient peoples and migrations without much more extensive and systematic sampling of ancient genetic material of a quality to permit the testing.

[Deirwha]

related follow up thought

As long as we are in this territory, I also do not understand why we test STR's as a beginning point. Why don't we test SNP's and then differentiate out from within defined clade with STR's and genetic distance. It has caused me no end of confusion and anxiety that taken as a whole I am at closer genetic distance to people who are not in my clade than to even the closest people in my clade. What purpose does it serve for me to see a genetic distance of 1 or 2 from people who are not in my clade at all, only to be told it is the result of convergent evolution and not to look at the fellow behind the curtain? Just some more musings on this process.

[john franks]

Some really helpful posts, thanks. I do feel that I have a better understanding of the whole subject now (how often can you say that about internet forums?)

I like cacio's reminder that the SNPs that have been chosen to trace the movement and whereabouts of people are just a few of many such changes. A lot of things flow from an understanding of this.

John

[MMaddi]

As long as we are in this territory, I also do not understand why we test STR's as a beginning point. Why don't we test SNP's and then differentiate out from within defined clade with STR's and genetic distance. It has caused me no end of confusion and anxiety that taken as a whole I am at closer genetic distance to people who are not in my clade than to even the closest people in my clade. What purpose does it serve for me to see a genetic distance of 1 or 2 from people who are not in my clade at all, only to be told it is the result of convergent evolution and not to look at the fellow behind the curtain? Just some more musings on this process.

The key to answering your question is that most of FTDNA's customers and those of other genetic genealogy testing companies are interested in genealogy. They may or may not have an interest in population genetics and the history, ancient and modern, of the nations and ethnic groups of their ancestors.

Since STR testing is most useful to establishing connections within a genealogical timeframe, that's what FTDNA and the other companies feature in their services. I think it's useful that FTDNA seems to have decided over the course of the last year to offer much more in the way of SNP testing. They also have ventured into SNP hunting, as seen in the L series SNPs.

[Deirwha]

thanks

that is a perfectly understandable answer and probably says a lot as to why I sometimes feel like my interests and questions are not understood. The frame of reference is different. I have little prospect of establishing a genealogical record on my father's side for the simple reasons there is a 70 year gap in the records caused by circumstances. I could make leaps and have kinda hoped that the DNA would help me bridge the gap, but thus far the gap has only widened because the Dyers who may be connected have either not taken the tests or have chosen to keep them private. I have had a long term personal interest, going back to junior high, in anthropology and that always returns as my focus. Hence I am more interested in tribe than establishing the "chain of custody" of my particular family line.