tag:blogger.com,1999:blog-7785493.post8709998058629965585..comments2024-01-04T04:11:55.717+02:00Comments on Dienekes’ Anthropology Blog: How Y-STR variance accumulates: a comment on Zhivotovsky, Underhill and Feldman (2006)Dienekeshttp://www.blogger.com/profile/02082684850093948970noreply@blogger.comBlogger34125tag:blogger.com,1999:blog-7785493.post-62083749374955276842013-03-18T15:44:32.258+02:002013-03-18T15:44:32.258+02:00I'm 13/12 at DYS19 and DYS388, respectively; t...I'm 13/12 at DYS19 and DYS388, respectively; the Greek Modal, and my highest 12 marker match frequency at FTDNA is for Greece at 3.7%. My paternal line is from northern Germany, near Hamburg.Levitylabhttps://www.blogger.com/profile/05374812313257748146noreply@blogger.comtag:blogger.com,1999:blog-7785493.post-53317968297370670312012-03-31T16:51:30.905+03:002012-03-31T16:51:30.905+03:00Dienkes, your calculations are likely to be more a...Dienkes, your calculations are likely to be more accurate than those using the Zhiv method. Based on the archaeological evidence, it is only from the Bronze Age / Late Neolithic that population demogrpaphy began to really grow, thus 'shape' the overall molecular profile of modern Europe.Robhttps://www.blogger.com/profile/07166839601638241857noreply@blogger.comtag:blogger.com,1999:blog-7785493.post-4974101779779759012009-01-11T13:56:00.000+02:002009-01-11T13:56:00.000+02:00no need for calibration or estimation of the germl...no need for calibration or estimation of the germline mutation rate ever.<BR/>if 1000 pairs of father/son observed and the mutation rate for an STR was found to be .002 then this is it folks.reletomphttps://www.blogger.com/profile/09584537578924459506noreply@blogger.comtag:blogger.com,1999:blog-7785493.post-71259566762178207522008-08-23T15:38:00.000+03:002008-08-23T15:38:00.000+03:00First, a couple of comments on ASD/VAR as methods ...First, a couple of comments on ASD/VAR as methods of counting mutations. I no longer use these techniques, I select DYS Loci where 95% of the allele values are at the modal and its two side values. This is, in part, to accommodate multi-step mutations which have been estimated as up to 5% of all mutations. Using the ASD/VAR squared differences counting technique significantly overcounts mutations. I also, currently, try not to use Palindromic dys loci, e.g. 385a,b due to the way they mutate occasionally.<BR/><BR/>I will now show that the Chandler rates do not predict known dates for sets of data. The ZUL rates appear to do much better, and at the present time I cannot explain why. I have studied your ZUL analysis and am hard-pressed to identify any errors in analysis. All I can say is I don't think it is the right model for mutations???<BR/><BR/>My approach is simple, I use a modified form of the Slatkin/Goldstein equation: TMRCA = (1/N)(1/#DL) X ( Sum over DL (#mu/musubr). Where N = number of Y chromosomes; #DL = number of dys loci; #mu = number of mutations at a dys loci; musubr = mutations rate of the dys loci.<BR/><BR/>The Kerchner data set has a mutation at 390, 439, 449(2), GATA H4, 576, CDYa, CDYb. (note some of these dys loci do not meet my criterion and I estimated the mutation rates using ASD, even though I think they're off. Fortunately, as we will show these dys loci contribute only a few years to the calculation). The ZUL mutation rate values for these dys loci (all in 10^-5): 4.17, 3.75, 12.67,4.17,9.75,16.2,16.2. Using the ten Kerchner entries and 37 dys loci, I compute a TMRCA of 305 years. Which appears to identify Adam, not Frederick as the founder. Using Chandlers rates and a 30 year gen to year conversion I get 109 years which is way too short!!!<BR/><BR/>For the Ian Cam analysis, all genetic descendants of the Clan Gregor founder who was born about 1300 AD, I only use the first 12 dys loci. I find mutations at the following dys loci of the 58 chromosomes I evaluated: 390, 1; 19, 1; 385a 1; 385b 2; 439 4; 389i 2;389ii 4. I omitted 3 mutations that are of close relatives: two Stirlings at 385b, and two 389i and 389ii who are close cousins. To the best of my knowledge that gave me unique transmission events.<BR/><BR/>Again the ZUL rates for these mutatins are: 390 4.17;19 1.17; 385a 1.67; 385b 4.90; 439 3.75; 389i 1.83; 389ii 4.67. Plugging these values in the equation with 12 dys loci and 58 entries I get a TMRCA = 1273 AD. This is within 25 years of the date of "reputed" founder and in fact could be his father? Chandler rates estimate 1641 as the founders date.<BR/><BR/>For whatever reason The ZUL rates provide me with precision and accuracy in these calculations. All the other calculations of TMRCA's for s21, 116+, Tarins Iberian data sets yield estimates consistent with traditional estimates!! (note in the Kerchner calculation the year contribution of CDYa and b is about 25 years total in the ZUL calculation)<BR/><BR/>If I have one regret here, it is that I cannot identify the error in logic in the VAR/Chandlers rate approach. I just believe it gives incorrect answers. BobMcGhttps://www.blogger.com/profile/03459589185170647441noreply@blogger.comtag:blogger.com,1999:blog-7785493.post-74107458801827185402008-08-18T14:23:00.000+03:002008-08-18T14:23:00.000+03:00I would still like to explore the Poisson distribu...I would still like to explore the Poisson distribution if I may and I thank you for your answer. It would seem the first questions I raised are consistent with your modelling? The Poisson distribution is a special case of the H/T binomial distribution when the probability of one of the events is rare. This is certainly our case since, over 37 dys loci, the P of no mutation is of the order .997 to .998? If I observe the properties of the allele histogram distribution of different dys loci I observe that generally it is more normal than uniform and it is usually skewed high or low. An exception is CDYa and b, which are approaching a uniform distribution, which is an expected result of a binomial process? So many of the distributions of dys loci do not meet a uniform dist. but appear to more fit a normal? (A good set of data to observe these distributions is the R.L.Tarin, Jr. data sets, on - line, at world families network.). So why are most dys loci not approaching a uniform distribution, which I believe should be the expected result??<BR/><BR/>Another general question about this process is: Are the numbers of mutations at a dys loci constrained?? One apparent constraint is the maximum allele length attainalble. It appears it is a number less than 50? Even more interesting is the question are the numbers of mutations at each dys loci, relative to other dys loci constrained? Goldstein and Stumpf's paper appears to say yes.(Science, 2 March 2001,vol. 291.) They argue, see eqn. 4, that any dys loci can be used to estimate TMRCA, whereas in practice, several are used and then averaged. This equation suggests that there is a balance between the number of mutations any one dys loci can have relative to the other dys loci. Do you have any idea what this constraint implies, beyond what I have said here???<BR/><BR/>Thank you for your thoughts. Next I plan to show how the ZUL rates, not the Chandler rates, predict the correct TMRCA for the Kerchner and Clan Gregor data sets.McGhttps://www.blogger.com/profile/03459589185170647441noreply@blogger.comtag:blogger.com,1999:blog-7785493.post-89946270670310236652008-08-17T23:12:00.000+03:002008-08-17T23:12:00.000+03:00the first major assumption in the analysis is that...<I>the first major assumption in the analysis is that the mean number of sons produced by a Y chromosome is 1. </I><BR/><BR/>The mean number of sons is 1 in a constant-sized population.<BR/><BR/>However, the mean number of sons _given that the line has survived to the present_ is not 1. These are the men we are interested in.<BR/><BR/>For example, in the beginning generation, each man has one son of average. But, over men who do have at least one son, the average number of sons is ~1.58.<BR/><BR/>And, if we are actually interested in men who are also MRCAs, then in the first generation they have ~2.39 sons on average (because they must have at least 2 sons if they are MRCAs).<BR/><BR/>So, while men have 1 son on average, if you look at the family trees of the men who have patrilineal descendants in the present, you will not see 1 son on average, but more.<BR/><BR/>Moreover, it is the case that people are usually interested in the big family trees, because these are more noticeable. Thus, if you look at men who don't just have patrilineal descendants today but a lot of them, you will find even a higher average number of sons.Dienekeshttps://www.blogger.com/profile/02082684850093948970noreply@blogger.comtag:blogger.com,1999:blog-7785493.post-34651325410622080932008-08-17T16:23:00.000+03:002008-08-17T16:23:00.000+03:00Addendum to my previous comment: m = 1 is the lea...Addendum to my previous comment: m = 1 is the least m can be if the population being studied is the successful Y chromosomes that reproduce - by definition. For m less than 1 and approaching zero, we have to include a much larger Y Chromosome population???McGhttps://www.blogger.com/profile/03459589185170647441noreply@blogger.comtag:blogger.com,1999:blog-7785493.post-89308983821979147772008-08-17T16:03:00.000+03:002008-08-17T16:03:00.000+03:00I would like to begin a detailed discussion of the...I would like to begin a detailed discussion of the ZUL derivation and especially its assumptions. I'm not sure if this is the right forum - but I don't know where else to go. Some of my comments may appear trivial and can be immediately dismissed, but I want to make sure that ZUL's model is correct, let alone his effective evolutionary rate.<BR/><BR/>First, Population. The population he models is a segment of the total population, the Y chromosome population that produces new Y chromosomes which may or may not have a mutation. This excludes all females, all-nonreproductive y chromosomes ( for any reason, e.g. all girls) etc. So, when we're talking about migrations of Gypsies, Maoris etc. We are only really discussing the Y chromosomes that produce new Y chormosomes.<BR/><BR/>Given these intrinsic assumptions, the first major assumption in the analysis is that the mean number of sons produced by a Y chromosome is 1. I can present two examples where this is not true. Now, I'm not blind to statistical assumptions and the fact that you are working with 10K Patriarchs and my examples may just be extremes within that population, but here goes anyway. First, the well documented Kerchner family which the Chandler rates don't model. I estimate that m = 1.6+ for this population. Second, my own heritage which I have good documentation for back to 1650 AD. I am the 11th generation and the average number of sons born down my line is 2.2+.<BR/><BR/>We may be looking at a little bit of the "Genghis Khan" phenomena in which some 5 to 6% of the asian population is attributable to him??? I don't know? The point is what is the "scientific" basis for assuming the Poisson process with a mean of 1 models Y Chromsome reproduction down a male line???McGhttps://www.blogger.com/profile/03459589185170647441noreply@blogger.comtag:blogger.com,1999:blog-7785493.post-80277386084628243262008-08-07T12:41:00.000+03:002008-08-07T12:41:00.000+03:00Actually a few linguists have recently suggested t...<I>Actually a few linguists have recently suggested that the Uralic expansion may be as recent as 2000 BC, which would agree with your dates very nicely.</I><BR/><BR/>That is very interesting, this is a related <A HREF="http://www.kotikielenseura.fi/virittaja/hakemistot/jutut/kallio1_2006.html" REL="nofollow">link</A> that someone sent me recently.<BR/><BR/><I>I hope you'll publish these soon.</I><BR/><BR/>I don't have plans to publish my results. I will, however, release my source code once I've tidied it up a bit and wrote some basic instructions, so that others can check its accuracy and repeat my calculations for any set of parameters they choose.Dienekeshttps://www.blogger.com/profile/02082684850093948970noreply@blogger.comtag:blogger.com,1999:blog-7785493.post-77104079355994195872008-08-07T10:15:00.000+03:002008-08-07T10:15:00.000+03:00Very intereting results, although I still need to ...Very intereting results, although I still need to think about your analysis in a bit more detail before I dare to say whether I agree with you. I hope you'll publish these soon.<BR/><BR/>I just wanted to comment on your recalculation of our dates for the Baltic region (Lappalainen et al. 2008), and the historical connections of N3. Actually a few linguists have recently suggested that the Uralic expansion may be as recent as 2000 BC, which would agree with your dates very nicely. It would indeed be quite intereting if BOTH linguistic and genetic dates become independently updated...Tuuli Lappalainenhttps://www.blogger.com/profile/02116822415452050812noreply@blogger.comtag:blogger.com,1999:blog-7785493.post-67734984437234353152008-08-02T23:25:00.000+03:002008-08-02T23:25:00.000+03:001st question: I believe that the convergence is n...1st question: I believe that the convergence is not to a man but to a haplotype. The Jefferson case illustrated that. In the Kerchner case, I estimate that Adam is the founder, not Frederick. I believe that Frederick may have been an only son, but he had three sons I believe. There no 10 patrilineal descendants. Google Charles Kerchner and you'll find his site and analysis all laid very carefully. Please note the distinctin between Unique Transmission events and transmittals.<BR/><BR/>2. In Clan Gregor, there are currently over 60 "genetic" MacGregors, descendants of the founder. The clan chieftain is hereditary and doesn't necessarily depict the actual number of scions. I think there is only one first descendant but I'll check and correct if I'm wrong.McGhttps://www.blogger.com/profile/03459589185170647441noreply@blogger.comtag:blogger.com,1999:blog-7785493.post-60679353868357953082008-08-02T21:46:00.000+03:002008-08-02T21:46:00.000+03:00>> I find that for short term family relatio...>> I find that for short term family relationships it works well, e.g. Kerchner family (350 years old), Clan Gregor (1300AD).<BR/><BR/>Two questions: are the founders of these clans also the MRCAs of their descendants. In other words, did they both have at least two sons with patrilineal descendants.<BR/><BR/>Also, how many patrilineal descendants are there (if such estimates exist).Dienekeshttps://www.blogger.com/profile/02082684850093948970noreply@blogger.comtag:blogger.com,1999:blog-7785493.post-81434543166415272942008-08-02T21:40:00.000+03:002008-08-02T21:40:00.000+03:00I have been studying TMRCA's using Zhivs equation....I have been studying TMRCA's using Zhivs equation. I find that for short term family relationships it works well, e.g. Kerchner family (350 years old), Clan Gregor (1300AD). I have also estimated TMRCA's for older sets of data and generally find my dates to be 2 to 3X Ken Nordtvedts who uses Chandlers rates. Note that Chandlers estimates uses many haplogroups, which I believe is incorrect. It appears that modal values and rates at some dy loci change over time. For a long time I used ASD to compute TMRCA's. I now feel it overcounts mutations at a dys loci by as much as a factor of 3 over simpler counting algorithms, where multi-step mutations are taken into account. I make no assumptions about population growth. I have not closely studied, I read it, your blog on this subject, I have found in the past that the assumptions are critical. I would like your first impression to my commentsMcGhttps://www.blogger.com/profile/03459589185170647441noreply@blogger.comtag:blogger.com,1999:blog-7785493.post-52729438676013064862008-08-01T13:51:00.000+03:002008-08-01T13:51:00.000+03:00J2b2-M241Aromuns (Albania, Macedonia, Romania) ......J2b2-M241<BR/><BR/>Aromuns (Albania, Macedonia, Romania) ..... 3.5 (1500BC) // E Bosch et al 2006<BR/>Europe .... 3.2 (1200BC) // ftdna, ysearch<BR/>India ..... 5.7 (3700BC) // Sengupta et al 2006<BR/>India ..... 4.9 (2900BC) // Jobling et al. 2006<BR/>Nepal ..... 4.2 (2200BC) // Jobling et al. 2006<BR/>South Asia (India, Malay, Nepal, Sri Lanka, Afghan) .... 5.6 (3600BC) // Jobling et al. 2006<BR/><BR/>* Klyesov's formula + Regular formulaUnknownhttps://www.blogger.com/profile/12546979477909894800noreply@blogger.comtag:blogger.com,1999:blog-7785493.post-50260209046694408042008-07-28T10:37:00.000+03:002008-07-28T10:37:00.000+03:00Exactly. Haplogroups distribution shows that thei...Exactly. Haplogroups distribution shows that their members have been moving around the earth considersbly. I maintain it's reasonable to assume such movement goes back, even beyond the evolution of our modern haplogroups.terrythttps://www.blogger.com/profile/17327062321100035888noreply@blogger.comtag:blogger.com,1999:blog-7785493.post-77947992490443620072008-07-27T06:24:00.000+03:002008-07-27T06:24:00.000+03:00terryt said,"'If we allow a single haplogroup to h...terryt said,<BR/><BR/>"'If we allow a single haplogroup to have evolved in two widely separated, genetically distinct ... populations.' As Maju says, that's impossible. However descendants of the common ancestor may have become widely spread before becoming sort of fixed in two separate populations, which is what appears to have happened in the examples mentioned here."<BR/><BR/>It is most definitely not impossible; in fact, it is exactly what Dienekes and others have argued for, that a particular haplogroup could have diversified and increased its proportion of the population in two or more separate regions at two or more different times. Notice that I used the word "evolved" (i.e. change and development; diversification) rather than "originated" (i.e. occurrence of the original mutation in the haplogroup's patriarch/founder; first appearance of the relevant SNP).<BR/><BR/>If we find evidence for this sort of scenario in the history of a particular haplogroup at the current level of phylogenetic resolution, then it is imperative that more detailed research be done to determine markers for subclades that developed in each of the distinct populations among which this haplogroup has become common, so that belonging to a particular subclade of that haplogroup will actually have some significance in relation to the overall genetic/ethnic background of an individual testee.Ebizurhttps://www.blogger.com/profile/16925110639823856429noreply@blogger.comtag:blogger.com,1999:blog-7785493.post-75104310189192831762008-07-27T04:48:00.000+03:002008-07-27T04:48:00.000+03:00"If we allow a single haplogroup to have evolved i..."If we allow a single haplogroup to have evolved in two widely separated, genetically distinct ... populations". As Maju says, that's impossible. However descendants of the common ancestor may have become widely spread before becoming sort of fixed in two separate populations, which is what appears to have happened in the examples mentioned here. <BR/><BR/>"the discriminatory power, or scientific significance, of the 'haplogroup R1b' designation is diminished, because the majority of the DNA of the average member of these populations (the autosomes) could have evolved independently and may be greatly dissimilar to each other". I have maintained that position all along. As Maju also says, "all ethnicities have their own salad of haplogroups" so a single haplogroup cannot be used as a marker for a particular ethnicity. Human groups have been mixing for as long as we have historical records (with occasional episodes of genocide) and probably were doing so long before we have these records.terrythttps://www.blogger.com/profile/17327062321100035888noreply@blogger.comtag:blogger.com,1999:blog-7785493.post-28442709049018946102008-07-26T21:38:00.000+03:002008-07-26T21:38:00.000+03:00The shared presence of a haplogroup in a large per...<I>The shared presence of a haplogroup in a large percentage of the members of any two or more populations is most significant if it indicates a shared period of evolution (both genetic and cultural) of those populations. Otherwise, haplogroups are really quite meaningless except on an individual scale for the purpose of genealogy (and even then, they can only help one find genealogical matches in the direct paternal or direct maternal line). </I><BR/><BR/>It depends. It specially depends on timelines. It's not the same if the haplogroup is 500, 5000 or 50,000 years old. The degree of cultural connection would obviously be much different. Also, specially for Paleolithic events, the likehood of some "random" founder effect or fixation via drift becomes much more likely, making the cultural connection even less clear.Majuhttps://www.blogger.com/profile/12369840391933337204noreply@blogger.comtag:blogger.com,1999:blog-7785493.post-43703481146678921242008-07-26T18:27:00.000+03:002008-07-26T18:27:00.000+03:00I started this last bit of discussion because Dien...I started this last bit of discussion because Dienekes had estimated the age of STR variation of J-M241 in the data of Pericic et al. 2005 (without Kosovars) on Balkan populations (I think they were mostly populations of the former Yugoslavia) to be 800 AD, while estimating the age of STR variation of J-M241 in the Indian data of Sengupta et al. 2006 to be 3,300 BC. These dates could be taken to suggest that haplogroup J2b1-M241 actually evolved among the ancestors of the Indians, and the presence of this haplogroup in a large percentage of modern males of populations in the Balkans does not really indicate anything significant about the ancient ancestry or genetic affinities of these Balkan populations (i.e., the shared presence of J-M241 between the Balkans and India does not indicate that there is any genetic connection between these two regions except maybe one male ancestor who very recently introduced J-M241 into the Balkans).<BR/><BR/>The shared presence of a haplogroup in a large percentage of the members of any two or more populations is most significant if it indicates a shared period of evolution (both genetic and cultural) of those populations. Otherwise, haplogroups are really quite meaningless except on an individual scale for the purpose of genealogy (and even then, they can only help one find genealogical matches in the direct paternal or direct maternal line).Ebizurhttps://www.blogger.com/profile/16925110639823856429noreply@blogger.comtag:blogger.com,1999:blog-7785493.post-53808028729406185632008-07-26T16:29:00.000+03:002008-07-26T16:29:00.000+03:00I don't understand well how this last bit of discu...I don't understand well how this last bit of discussion arose but it's clear that it's statistically nigh-on-impossible that an SNP arose twice in human biological history. When there is more than one defining SNP, then the situation is even more clear. Hence R1b means one single common male ancestor via purely paternal line for all carriers of that haplogroup. <BR/><BR/>But that doesn't mean that, after the clade arose, different descendants could not have got different ethnic identities, lived in different geographic locations or whatever. A haplogroup is not the same as ethnicity. In theory there is no objection to some R1b being Basque and some R1b being Indo-European, no matter they share the same common ancestor. In fact all ethnicities have their own salad of haplogroups and related haplos are found among very different ethnic groups.Majuhttps://www.blogger.com/profile/12369840391933337204noreply@blogger.comtag:blogger.com,1999:blog-7785493.post-24079926678452951642008-07-26T08:44:00.000+03:002008-07-26T08:44:00.000+03:00terryt asked,"So what precisely do you see to be t...terryt asked,<BR/><BR/>"So what precisely do you see to be the problem with those ideas?"<BR/><BR/>I just think that it would negate the usefulness of Y-DNA and mtDNA testing for the purpose of identifying one's "deep ancestry." If we allow a single haplogroup to have evolved in two widely separated, genetically distinct (in terms of their autosomes) populations, then it greatly reduces the significance of the determination that an individual belongs to that haplogroup.<BR/><BR/>For example, if R1b is allowed to be both Basque and Indo-European in origin, then the discriminatory power, or scientific significance, of the "haplogroup R1b" designation is diminished, because the majority of the DNA of the average member of these populations (the autosomes) could have evolved independently and may be greatly dissimilar to each other. A haplogroup is much more meaningful if it indicates a shared evolutionary origin, i.e. an origin in a particular endogamous group that has developed autosomal genetic distinctiveness to match the Y-DNA or mtDNA "signature."<BR/><BR/>Besides, allowing so many separate expansion events (e.g. hypothesizing that haplogroup J2b1-M241 expanded to become a major haplogroup in at least two separate populations in two widely separated regions) makes the possibility of "hidden archaic admixture" that much more plausible, because we can't know what other haplogroup(s) might have been lurking in one area or the other when J2b1-M241 managed to gain a foothold and expand there.Ebizurhttps://www.blogger.com/profile/16925110639823856429noreply@blogger.comtag:blogger.com,1999:blog-7785493.post-38673728659065430052008-07-26T00:26:00.000+03:002008-07-26T00:26:00.000+03:00The ultimate origin of J-M12 is usually placed in ...The ultimate origin of J-M12 is usually placed in the Near East, however there is little of J-M12 there. I wouldn't speculate on this point.<BR/><BR/><I>aren't we opening the floodgates to ideas like "R1b in Iberia is Basque, but R1b in Armenia is Indo-European" </I><BR/><BR/>A haplogroup need not have an expanded in one place. J1 for example multiplied among Arabians and Northeast Caucasian speakers.Dienekeshttps://www.blogger.com/profile/02082684850093948970noreply@blogger.comtag:blogger.com,1999:blog-7785493.post-91291159349340680552008-07-26T00:04:00.000+03:002008-07-26T00:04:00.000+03:00dienekes wrote,"This is the dating of J-M241 _in t...dienekes wrote,<BR/><BR/>"This is the dating of J-M241 _in the studied populations_ and not the dating of J-M241 in general.<BR/><BR/>The MRCA of J-M241 in the Balkans isn't the same as the MRCA of J-M241 in India. Both of them are descended from a common ancestor (the M241 guy), but I am not dating _him_."<BR/><BR/>Then you are presuming that J-M241 had produced several descendants, at least one of which went to northern Greece (and Albania, etc.) and produced prolific progeny in that region, and at least one other of which went to northern India/Nepal and whose descendants experienced a separate period of population growth there. Where is the location of haplogroup J-M241's ultimate origin? Is the origin of J-M241 as a whole even determinable if one allows the sort of scenario that you are describing?<BR/><BR/>If we start accepting the idea that a clade could have expanded separately in two distant locations, aren't we opening the floodgates to ideas like "R1b in Iberia is Basque, but R1b in Armenia is Indo-European" and "Neanderthal DNA did introgress into European Homo sapiens sapiens, but archaic Y-DNA and mtDNA lineages have been lost due to genetic drift"?Ebizurhttps://www.blogger.com/profile/16925110639823856429noreply@blogger.comtag:blogger.com,1999:blog-7785493.post-38341020790926781782008-07-25T23:42:00.000+03:002008-07-25T23:42:00.000+03:00This is the dating of J-M241 _in the studied popul...This is the dating of J-M241 _in the studied populations_ and not the dating of J-M241 in general.<BR/><BR/>The MRCA of J-M241 in the Balkans isn't the same as the MRCA of J-M241 in India. Both of them are descended from a common ancestor (the M241 guy), but I am not dating _him_.Dienekeshttps://www.blogger.com/profile/02082684850093948970noreply@blogger.comtag:blogger.com,1999:blog-7785493.post-4112842607668306892008-07-25T23:37:00.000+03:002008-07-25T23:37:00.000+03:00Dienekes wrote,"The young ages of J-M12 and J-M241...Dienekes wrote,<BR/><BR/>"The young ages of J-M12 and J-M241 also explain the striking inverse correlation between it and J-M410, which makes sense if it expanded later. A fairly late expansion also explains its under-representation in Southern Italy and Anatolia: it appears to be a rather young and "Epirotic" clade that was too late in coming to significantly participate in the historical Greek colonization."<BR/><BR/>I've enjoyed reading about your conversions from Zhivotovsky's hypothetical "evolutionary mutation rates" to observed "pedigree mutation rates," and I don't mean to sound too critical of your estimates, but how would you explain the significant presence of haplogroup J-M241 in India and Nepal if it is an "Epirotic" clade that has expanded so recently? Something having to do with Alexander's invasion or the Yavanas? But Alexander was a Macedonian, and "Yavana," in its strictest sense, is supposed to refer to Ionians. Within Europe, haplogroup J-M241 is presently found most commonly in Albania, is it not? What historical movement could have brought haplogroup J-M241 to Albania and Nepal in such great quantities at a date so recent as that which you have proposed for this clade's expansion?Ebizurhttps://www.blogger.com/profile/16925110639823856429noreply@blogger.com