On Wednesday, June 8, 2016 at 11:24:08 AM UTC-7, Thomas Milton Tinney, Sr. wrote: > > REPLY: No, this is not a reply, it is just another misapplied Google search result. > --------------------------- > ADDENDUM: (pdf)from Cornell University Library; > Open access to 1,154,307 e-prints in Physics, Mathematics, Computer > Science, Quantitative Biology, Quantitative Finance and Statistics > https://arxiv.org/pdf/1602.08183.pdf Except is is not from Cornell University Library, except that they host a server that lets anyone dump a paper there ArXiv.org, intended to allow a crowd-sourced pre-review of a paper the author intends to submit for publication, but has not yet. The papers on ArXiv are not (or have yet to be) peer reviewed when they are deposited there, and that is the case with this paper. The author did give an oral presentation with this title at the 2012 Proceedings of the 8th international conference on Bioinformatics Research and Applications, but again, no peer review (and a long lag time after the presentation before this paper was put forward, so there might be some problems there). > > Non-Identafible Pedigrees and a Bayesian Solution dated 26 Feb 2016 > > . . . > > Abstract. > > "Some methods aim to correct or test for relationships or to reconstruct > > the pedigree, or family tree. We show that these methods cannot resolve > > ties for correct relationships due to identifiability of the pedigree > > likelihood which is the probability of inheriting the data under the etc. > > In suggesting a potential solution, mention is made that: > > "As yet, all these details are an open problem." Did you notice the paper is talking about forensics - determining the precise relationship of a DNA sample of an unknown person left at a crime scene to a single individual in a database, if they are more distant that siblings or parent-child when looking at multiple unlinked loci (i.e. a grandfather, and uncle and a half-brother all have the potential to show similar degrees of shared markers). This is not something that genetic genealogy has ever promised or aimed at accomplishing. Further, the fact that you are reading a paper saying that this is something DNA analysis cannot do tells you it is not a dirty little secret, but recognized as something that DNA analysis cannot do. Every professional American genealogist who has worked with the record knows that the 1850 US census cannot be used to determine the precise relationship between two individuals because it does not contain the necessary detail. It would be ridiculous to claim that all censuses are worthless because of this, but that is essentially what you are doing - suggesting that because there is a question it cannot answer, and practicing professionals know it cannot answer, and suggesting it negates DNA not only for addressing that question, but all applications of DNA to any question. It is absolutely ludicrous. > > That "The origins, effects and frequency of mutations has already > > been fully integrated into our understanding" falls into the category > > of comic DNA fantasy. On what basis do you conclude this. To be able to reach this conclusion you you must either be an expert on DNA, or comic fantasy. You have demonstrated that DNA is not a direction in which your expertise lies. So far, excluding all of the red herrings and straw-man fallacies, your entire argument has boiled down to, "Because DNA!" and that has proven underwhelming in its power of persuasion. taf

On Wednesday, June 8, 2016 at 12:34:37 PM UTC-7, taf wrote: > On Wednesday, June 8, 2016 at 11:24:08 AM UTC-7, Thomas Milton Tinney, Sr. wrote: > > > > REPLY: > > No, this is not a reply, it is just another misapplied Google search result. > > > > --------------------------- > > ADDENDUM: (pdf)from Cornell University Library; > > Open access to 1,154,307 e-prints in Physics, Mathematics, Computer > > Science, Quantitative Biology, Quantitative Finance and Statistics > > https://arxiv.org/pdf/1602.08183.pdf > > > Except is is not from Cornell University Library, except that they host a server that lets anyone dump a paper there ArXiv.org, intended to allow a crowd-sourced pre-review of a paper the author intends to submit for publication, but has not yet. The papers on ArXiv are not (or have yet to be) peer reviewed when they are deposited there, and that is the case with this paper. > > The author did give an oral presentation with this title at the 2012 Proceedings of the 8th international conference on Bioinformatics Research and Applications, but again, no peer review (and a long lag time after the presentation before this paper was put forward, so there might be some problems there). > > > > > Non-Identafible Pedigrees and a Bayesian Solution dated 26 Feb 2016 > > > . . . > > > Abstract. > > > "Some methods aim to correct or test for relationships or to reconstruct > > > the pedigree, or family tree. We show that these methods cannot resolve > > > ties for correct relationships due to identifiability of the pedigree > > > likelihood which is the probability of inheriting the data under the > > etc. > > > > In suggesting a potential solution, mention is made that: > > > "As yet, all these details are an open problem." > > Did you notice the paper is talking about forensics - determining the precise relationship of a DNA sample of an unknown person left at a crime scene to a single individual in a database, if they are more distant that siblings or parent-child when looking at multiple unlinked loci (i.e. a grandfather, and uncle and a half-brother all have the potential to show similar degrees of shared markers). This is not something that genetic genealogy has ever promised or aimed at accomplishing. > > Further, the fact that you are reading a paper saying that this is something DNA analysis cannot do tells you it is not a dirty little secret, but recognized as something that DNA analysis cannot do. Every professional American genealogist who has worked with the record knows that the 1850 US census cannot be used to determine the precise relationship between two individuals because it does not contain the necessary detail. It would be ridiculous to claim that all censuses are worthless because of this, but that is essentially what you are doing - suggesting that because there is a question it cannot answer, and practicing professionals know it cannot answer, and suggesting it negates DNA not only for addressing that question, but all applications of DNA to any question. It is absolutely ludicrous. > > > > That "The origins, effects and frequency of mutations has already > > > been fully integrated into our understanding" falls into the category > > > of comic DNA fantasy. > > On what basis do you conclude this. To be able to reach this conclusion you you must either be an expert on DNA, or comic fantasy. You have demonstrated that DNA is not a direction in which your expertise lies. > > So far, excluding all of the red herrings and straw-man fallacies, your entire argument has boiled down to, "Because DNA!" and that has proven underwhelming in its power of persuasion. > > taf ------------------------------------- REPLY: (1) Distortion of genealogical properties when the sample is very large . . . "We will soon enter an era in which it will become routine to analyze samples with hundreds of thousands if not millions of individuals. For these large sample sizes, the standard coalescent will no longer serve as an adequate model for evolution. The DTWF model is mathematically cumbersome to work with, which was one of the original motivations for adopting the coalescent for modern population genetics analyses. However, for these very large sample sizes, we will need to develop new mathematically and computationally tractable stochastic processes that better approximate realistic models of human population evolution, and under which we can efficiently compute genealogical quantities like we have been able to under the coalescent." http://www.pnas.org/content/111/6/2385.long (2) Fundamental limits on the accuracy of demographic inference based on the sample frequency spectrum Significance "Numerous empirical studies in population genetics have used a summary statistic called the sample frequency spectrum (SFS), which summarizes the information in a sample of DNA sequences. Despite their popularity, the accuracy of inference methods based on the SFS is difficult to characterize theoretically, and it is currently unknown how the estimation accuracy improves as more sites in the genome are used. Here, we establish information theoretic limits on the accuracy of all estimators that use the SFS to infer population size histories. We study the rate of convergence to the true answer as the amount of data increases, and obtain the surprising result that it is exponentially worse than known convergence rates for many classical estimation problems in statistics." http://www.pnas.org/content/112/25/7677 As I stated heretofore, promises, promises, promises, as larger databases become available, reveal more complexity in the systems proposed analysis, and for which "we will need to develop new mathematically and computationally tractable stochastic processes that better approximate realistic models of human population evolution". DNA is not the problem. As of 2015 and 2016 era papers, full public admission is made that models used in the past are going to be readjusted to "efficiently compute genealogical quantities". All this is bad news to the family history community. They have been led to believe that as the size of their genetic sample base increases, via $$$ transfers to various sampling companies, they will have further stunning improvements in pedigree accuracy. NOT SO! And so the cat is out of the bag, finally, finally! The scientific community needs the genealogy community to give it the parameters it needs to figure out future approximate estimates on what its independently (so called) derived assumptions and mathematical calculations are based. Dollars $$$ will thus continuously flow from "widows pockets", genealogists and family historians, since with every new future adjustment, past assessments will all be ancient history. This is an ideal economic and business model; a fabulous way to keep coffers full!

On Wednesday, June 8, 2016 at 10:00:21 PM UTC-7, Thomas Milton Tinney, Sr. wrote: > ------------------------------------- > REPLY: No, it is not a reply, just more irrelevant papers you don't understand. I tire of this game. This time, both papers had to do with coalescence date estimates from very large populations - time spans of <10s of thousands of years, and hence completely irrelevant, YET AGAIN. > DNA is not the problem. Wait, so we ave spent the last week with you complaining about DNA being a big problem, and now DNA isn't the problem? > As of 2015 and 2016 era papers, full public admission is made that models > used in the past are going to be readjusted to "efficiently compute > genealogical quantities". All this is bad news to the family history > community. Only if they are doing population genetics on the side, since that is what the papers were talking about, not genealogy. > They have been led to believe that as the size of their genetic sample > base increases, via $$$ transfers to various sampling companies, they will > have further stunning improvements in pedigree accuracy. Oranges, meet apples. Nothing whatsoever to do with the studies you are citing. > The scientific community needs the genealogy community to give it the > parameters it needs to figure out future approximate estimates on what > its independently (so called) derived assumptions and mathematical > calculations are based. It doesn't work this way - unless you want to endow a research program dedicated to satisfying your whims. Only then do you get to decide what scientists (the ones you employ) need to do. > Dollars $$$ will thus continuously flow from "widows pockets", Oh, the humanity! All those poor widows being provided with the exact service they desire, for the market cost! The government should put a stop to such blatant (gasp) capitalism. > genealogists and family historians, since with every new future > adjustment, past assessments will all be ancient history. This is an > ideal economic and business model; a fabulous way to keep coffers full! And while you are at it, let's condemn Apple. I mean, really, they are always coming out with a new phone and all of those desperate 30-somethings have to run right out and buy it even though they have a perfectly good phone already, the poor bastards. Isn't progress terrible! If you prefer to wait for the next better thing coming down the pike, hold onto your money until then. If you don't want to pay for an improved product, don't buy it. taf