I would agree with Wjhonson, except I would tend to want to put it in even stronger terms. I would say every ancestor passes persistent segments along to each of us. I also agree with John McCoy, except I tend to think his statement about 'statistical oddity' is there because nobody has studied it in depth. Or, nobody has figured out that it should be studied. The repercussion of our ignorance of the topic leads to problems such as my vendor only being able to capture half of my fourth and fifth cousins. Some say that not everybody passes along DNA due to the random nature of inheritance. I would tend to say that my 4th and 5th cousins are not being detected by the vendors not only because of the 7 cM limit, but also because the vendors are looking at the wrong thing. Vendors base relatedness on the size in cMs (some call this 'Genetic Distance), when they should be using cMs in order to determine TMRCA, and then base relatedness upon TMRCA. What seems to be persistent in my I1-M253 group are the triads that I find. However, determining the age of these small segments has been a problem. We are all taught that the Y-DNA SNPs are fairly stable, and that the Y-DNA haplotype groups can persist for thousands of years. Each Y-DNA haplotype group representing a very ancient family. Yet, nobody regularly teaches us that autosomal DNA SNPs may also persist for thousands of years. As I have mentioned before, the mother to child segments passed appear to follow an SNP to cMs ratio (SNPS/cMs) of 200. However, half of my data from GEDMatch has a ratio of less than 200. When I look at Ancient DNA reconstructions at GEDMatch, I find that nearly all of the Ancient DNA samples carry shared segments of less than 200. Which is to say, the ancient segments apparently, do not want to change. Since half of my 4th and 5th cousins are not being detected at FTDNA (due to the segment size theory), my thought process has evolved along the line of dating segments, rather than dating segment sizes. This has resulted in an improved phylogenetic tree than my previous thoughts on how to interpret Genetic Distance of autosomal segments. That leaves me with the statistical analysis that John McCoy mentioned. Namely, how robust is this thought, and how do I eliminate errors across haplotype groups?  - Dave Hamm RE: On 11/21/2018 18:37 PM, Betsy Shafer wrote: as someone who studied chemistry quite a bit . . . sometimes more than other times . . . i would say what you are looking at is 'stable' segments, perhaps we could say 'stable configurations' . . . of course we can't attribute any motivations or emotional aspects to these chains of atoms, but we can say that nature seems to make organisms 'want' to preserve and pass on traits . . . those traits that contribute to the 'fitness' (stability? ability to fend for itself? healthiness?) of an organism or a strain of organisms are more likely to endure as the organism (or maybe line of ancestors) procreates and endures through time . . . - b On 11/21/2018 7:13 PM, John McCoy wrote: > This might be the same sort of observation often termed "meiotic drive" in the genetics literature.  (And I prefer not to use terminology that attributes motives to genes!)  Yes, this sort of thing has been observed many times and at least sometimes seems to be something more than a statistical oddity.  Whether it can be invoked as the "cause" of any apparently persistent segment that we see in genetic genealogy is problematic, since we will rarely have a large enough sample to established statistical significance. > John McCoy(RealMac@aol.com) In a message dated 11/22/2018 Wjhonson writes: I would make your claim stronger, and say that persistent segments are in fact very common. It's just that you have no expectation of a persistent segment from any specific ancestor, but you certainly have some from some ancestors. I would say every person has some. > In a message dated 11/21/2018 3:23:23 PM PST Wesley Johnston writes: > > I just listened to the Economist Babbage podcast in which they discussed ways of dealing with malaria through mosquito genetics. One idea is to take what has been already found in nature -- what is called a "selfish gene" -- and have it spread across the malaria-bearing mosquito population (which is actually 3 different species of mosquito).A selfish gene is one that somehow aids and abets its own propagation to the next generation. These have apparently been recognized as a reality in nature for some time.We have had discussions on this list about autosomal DNA that has persisted over multiple generations in excess of what the assumption of 50-50 average recombination would lead us to expect -- but which we have found in the reality of the descendants of some ancestor much further back than would be expected by so many descendants sharing the DNA. It is not at all common, but it is there in some cases.I do not know enough about genetics to say whether the "selfish gene" observed in nature is connected to what we are seeing in some cases in the atDNA of descendants of some ancestors.