Announcement

Collapse
No announcement yet.

Eurogenes K13 and K15 4-Ancestors Oracle

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • 1798
    replied
    Maybe someone can explain this for me. What dna did this Yamnaya man's relatives bring to Ireland? He certainly did not bring any because he died in Russia.

    Population Irish K15
    North_Sea 34.43%
    Atlantic 31.38%
    Baltic 13.42%
    Eastern_Euro 7.33%
    West_Med 5.38%
    West_Asian 5.47%
    East_Med -*
    Red_Sea -*
    South_Asian 1.95%
    Southeast_Asian -*
    Siberian -*
    Amerindian 0.44%
    Oceanian -*
    Northeast_African 0.20%



    Yamnaya 10176
    North_Sea 17.98
    Atlantic 9.73
    Baltic 14.12
    Eastern_Euro 26.52
    West_Med 0
    West_Asian 21.1
    East_Med 0
    Red_Sea 0
    South_Asian 5.99
    Southeast_Asian 0
    Siberian 0
    Amerindian 4.45
    Oceanian 0.12
    Northeast_African 0
    Sub-Saharan 0
    Last edited by 1798; 24 April 2015, 04:25 AM.

    Leave a comment:


  • 1798
    replied
    Originally posted by 192971 View Post
    I did not say exactly that. They can reliably tell what averaged modern population has the most similar admixture to an individual having ancestry mostly from that one population. Also, they can reliably tell when the situation is nothing like the before mentioned. When two modern population components are to be chosen to represent the ancestry of an individual, the calculators start to be less reliable due to methodology, and ever more when more populations are added to calculation.

    In the end, no calculator can fix the problem that you get uneven shares of aDNA from your four grandparents, and that half of your parents' aDNA was randomly discarded in composition of your aDNA. So, there never will be a calculator which would totally agree with the admixtures of your (grand)parents. Family trees and genetics will show different results forever.
    The calculator got it 75% right for my dna.

    Leave a comment:


  • N21163
    replied
    Originally posted by 192971 View Post
    I did not say exactly that. They can reliably tell what averaged modern population has the most similar admixture to an individual having ancestry mostly from that one population. Also, they can reliably tell when the situation is nothing like the before mentioned. When two modern population components are to be chosen to represent the ancestry of an individual, the calculators start to be less reliable due to methodology, and ever more when more populations are added to calculation.

    In the end, no calculator can fix the problem that you get uneven shares of aDNA from your four grandparents, and that half of your parents' aDNA was randomly discarded in composition of your aDNA. So, there never will be a calculator which would totally agree with the admixtures of your (grand)parents. Family trees and genetics will show different results forever.
    Thank you for a detailed explanation.

    Leave a comment:


  • 1798
    replied
    Originally posted by robertalabama View Post
    So, I did the ancient DNA comparison on Gedmatch and had a 3cm segment in common with this Clovis-Anzick. I realize this segment is too short to be considered a "match," but does it mean anything at all ?
    http://www.y-str.org/2014/09/clovis-anzick-dna.html



    Also.
    http://news.nationalgeographic.com/n...rst-americans/
    "Comparison studies of the ancient DNA showed that it was similar to the genomes of ancient people living in Siberia and the ancestors of East Asians. The team also discovered a deep genetic affinity between the boy's genetic material and those of 52 Native American populations living in South America and Canada."

    Leave a comment:


  • 192971
    replied
    Originally posted by 1798 View Post
    So, you are saying these calculators are useless. Someone should create one that we can use.
    I did not say exactly that. They can reliably tell what averaged modern population has the most similar admixture to an individual having ancestry mostly from that one population. Also, they can reliably tell when the situation is nothing like the before mentioned. When two modern population components are to be chosen to represent the ancestry of an individual, the calculators start to be less reliable due to methodology, and ever more when more populations are added to calculation.

    In the end, no calculator can fix the problem that you get uneven shares of aDNA from your four grandparents, and that half of your parents' aDNA was randomly discarded in composition of your aDNA. So, there never will be a calculator which would totally agree with the admixtures of your (grand)parents. Family trees and genetics will show different results forever.
    Last edited by 192971; 22 April 2015, 01:21 AM.

    Leave a comment:


  • 1798
    replied
    Originally posted by 192971 View Post
    Just in case you guys do not know this...

    1. Your distance to yourself in a calculator would be zero.
    2. The theoretical maximum distance between an individual and a modern population would be 100.

    The distance value after @ sign is calculated as an Euclidean distance of two points in N-dimensional space, where N is the number of ancestral population components "found" when the admixture analyzer was developed. For example, modern population fitting to an individual's K15 results utilizes 15-dimensional orthogonal coordinate system.

    If more than one modern population is used in the fitting, they are admixed mathematically by (weighted) averaging of their ancestral population component percentages. That is how one of the points is determined. The other point is the individual's, of course.

    Euclidean distances are simple to calculate and obvious tools, but a problem is that they are abstract measures of something somewhat applicable here.
    So, you are saying these calculators are useless. Someone should create one that we can use.

    Leave a comment:


  • N21163
    replied
    Originally posted by 192971 View Post
    Just in case you guys do not know this...

    1. Your distance to yourself in a calculator would be zero.
    2. The theoretical maximum distance between an individual and a modern population would be 100.

    The distance value after @ sign is calculated as an Euclidean distance of two points in N-dimensional space, where N is the number of ancestral population components "found" when the admixture analyzer was developed. For example, modern population fitting to an individual's K15 results utilizes 15-dimensional orthogonal coordinate system.

    If more than one modern population is used in the fitting, they are admixed mathematically by (weighted) averaging of their ancestral population component percentages. That is how one of the points is determined. The other point is the individual's, of course.

    Euclidean distances are simple to calculate and obvious tools, but a problem is that they are abstract measures of something somewhat applicable here.
    I probably couldn't have explained it better. Thanks for posting this.

    Leave a comment:


  • robertalabama
    replied
    The fact that Clovis is closest to the modern Native American peoples says it all.
    So, I did the ancient DNA comparison on Gedmatch and had a 3cm segment in common with this Clovis-Anzick. I realize this segment is too short to be considered a "match," but does it mean anything at all ?

    Leave a comment:


  • 1798
    replied
    When Q1a3a crossed the Bering Strait there was no R of any kind in Siberia so Mal'ta R did not survive. The Samara HG R1b is a long way time away from Mal'ta. And furthermore, the NA were not black.

    Leave a comment:


  • 192971
    replied
    About distances in population fitting

    Just in case you guys do not know this...

    1. Your distance to yourself in a calculator would be zero.
    2. The theoretical maximum distance between an individual and a modern population would be 100.

    The distance value after @ sign is calculated as an Euclidean distance of two points in N-dimensional space, where N is the number of ancestral population components "found" when the admixture analyzer was developed. For example, modern population fitting to an individual's K15 results utilizes 15-dimensional orthogonal coordinate system.

    If more than one modern population is used in the fitting, they are admixed mathematically by (weighted) averaging of their ancestral population component percentages. That is how one of the points is determined. The other point is the individual's, of course.

    Euclidean distances are simple to calculate and obvious tools, but a problem is that they are abstract measures of something somewhat applicable here.

    Leave a comment:


  • 1798
    replied
    Originally posted by N21163 View Post
    I never said you did.



    That some modern Native American populations descend from the same people as the Clovis sample yes.



    No I'm very aware of what you try and do and you often get ahead of yourself with trying to oversimplify things.

    At this point the most that you could possibly deduce is that you and Hinxton-4 may have ancestry in common from the Iron Age...possibly...assuming all of the parameters of the calculator are correct.



    So you're still trying to state that La Brana is Irish??

    But wait a minute...does this mean that you and the Hinxton-4 sample both have North German ancestry as well??

    You: 1 Irish + Irish + Irish + North_German @ 4.162835

    Hinxton-4: 1 Irish + Irish + Irish + North_German @ 4.858814

    But how could that be?? Aren't you 100% Irish?? You've told us all numerous times that all of your ancestors came from Ireland...
    So is there work to be done on the calculator or is there more to your ancestry? Or both?? Wouldn't want to make too many assumptions too quickly there!


    I'll try and explain the admixture calculators again so that you can understand.
    There are a limited number of sample populations that are listed as having "average" percentages of a number of different groups.
    If the end number is higher, it is less likely that the sample will be from these populations.
    The high numbers are an issue for ancient samples, as you are comparing them to modern populations.



    Finally! Now that wasn't so hard was it? A best fit has limitations though. Further refining of the calculators is needed, this has been addressed numerous times.



    Adjust what exactly? The end number? It's an output number generated from the algorithm calculator. To change the end number the sample population "averages" would need to be altered, more populations would need to be added or a new set of parameters would need to be developed. I'm not sure how you know about these sorts of things but it's not like flicking a light switch on and off. It takes some time to develop.
    1.You contradict yourself.

    2.I knew that already.

    3.You underestimate me.

    4.I am not the only Irish,Welsh or Scottish person with some north German ancestry and they belong in other "Y" branches also.The north German links could be ancient like the Funnel Beaker Culture who built Dolmens and Passage Tombs like those that are found in Ireland. This Culture may have had the LP gene.

    Leave a comment:


  • N21163
    replied
    Originally posted by 1798 View Post
    I don't remember stating that Clovis had descendants.
    I never said you did.

    Originally posted by 1798 View Post
    The fact that Clovis is closest to the modern Native American peoples says it all.
    That some modern Native American populations descend from the same people as the Clovis sample yes.

    Originally posted by 1798 View Post
    I have already shown you that I am close to Hinxton 4 and my autosomal dna represents the modern Irish. What part do you not understand?
    No I'm very aware of what you try and do and you often get ahead of yourself with trying to oversimplify things.

    At this point the most that you could possibly deduce is that you and Hinxton-4 may have ancestry in common from the Iron Age...possibly...assuming all of the parameters of the calculator are correct.

    Originally posted by 1798 View Post
    That does not mean that it is wrong even though the end number is high.
    So you're still trying to state that La Brana is Irish??

    But wait a minute...does this mean that you and the Hinxton-4 sample both have North German ancestry as well??

    You: 1 Irish + Irish + Irish + North_German @ 4.162835

    Hinxton-4: 1 Irish + Irish + Irish + North_German @ 4.858814

    But how could that be?? Aren't you 100% Irish?? You've told us all numerous times that all of your ancestors came from Ireland...
    So is there work to be done on the calculator or is there more to your ancestry? Or both?? Wouldn't want to make too many assumptions too quickly there!


    I'll try and explain the admixture calculators again so that you can understand.
    There are a limited number of sample populations that are listed as having "average" percentages of a number of different groups.
    If the end number is higher, it is less likely that the sample will be from these populations.
    The high numbers are an issue for ancient samples, as you are comparing them to modern populations.

    Originally posted by 1798 View Post
    Lochsbour is closest to the modern Lithuanian and Swedish populations.The calculator picks the best fit population for every query.
    Finally! Now that wasn't so hard was it? A best fit has limitations though. Further refining of the calculators is needed, this has been addressed numerous times.

    Originally posted by 1798 View Post
    The designer can adjust it if he wishes to do so.
    Adjust what exactly? The end number? It's an output number generated from the algorithm calculator. To change the end number the sample population "averages" would need to be altered, more populations would need to be added or a new set of parameters would need to be developed. I'm not sure how you know about these sorts of things but it's not like flicking a light switch on and off. It takes some time to develop.
    Last edited by N21163; 21 April 2015, 07:29 AM.

    Leave a comment:


  • Subwoofer
    replied
    Originally posted by 1798 View Post
    Academics in Ireland don't know who exactly the Celts where or when they first came to Ireland. And then someone who is not from Ireland writes that they know the type of blood the Celts had.
    Blood of the Celts is Jean Manco's new book which comes out in the Autumn, possibly you could wait until then before reviewing it : )

    Probably it will be very good, Jean is an historian and is very considered with her research, plus (unlike some) she's not afraid to change her mind when new evidence presents itself.

    Leave a comment:


  • 1798
    replied
    I don't remember stating that Clovis had descendants. The fact that Clovis is closest to the modern Native American peoples says it all.

    I have already shown you that I am close to Hinxton 4 and my autosomal dna represents the modern Irish. What part do you not understand?

    Lochsbour is closest to the modern Lithuanian and Swedish populations. The calculator picks the best fit population for every query. That does not mean that it is wrong even though the end number is high. The designer can adjust it if he wishes to do so.

    Leave a comment:


  • N21163
    replied
    Originally posted by 1798 View Post
    Clovis F999919 K15
    Population
    North_Sea 0.95%
    Atlantic 0.14%
    Baltic 0.08%
    Eastern_Euro 3.52%
    West_Med -
    West_Asian -
    East_Med -
    Red_Sea -
    South_Asian -
    Southeast_Asian 1.63%
    Siberian 4.58%
    Amerindian 88.51%
    Oceanian 0.34%
    Northeast_African 0.26%
    Sub-Saharan -

    1 Anzick-1 + Pima + Pima + Pima @ 2.099648

    Clovis was a NA. Is this not correct? K15 looks okay to me but it may need a little tweaking.
    I get the impression you do not care to read through long posts, but anyway.

    I don't recall mentioning the Clovis sample before, did you mention it and I didn't respond?
    Didn't think so.

    The Clovis sample is 12,500 years old and it has been established that some Native American groups descend from the peoples that this sample belongs to. No one directly descends from this sample as it belongs to an infant male.
    Agreed so far?

    I thank you for including the 4 population matches from the gedmatch results.
    "Anzick-1" refers to the "Clovis" sample itself, and anyone can search this on wikipedia: http://en.wikipedia.org/wiki/Anzick-1

    The sample was named after the family who resided on the land, in Montana USA, where it was discovered. Sarah Anzick (of the Anzick family) is a Molecular Biologist who decoded the sample and assisted in reburial of it's remains: https://www.sciencenews.org/article/...-ancient-bones

    Looking at the 4 population approximation you porvided:

    1 Anzick-1 + Pima + Pima + Pima @ 2.099648

    the gedmatch calculator is stating the sample descends from itself (??) (in the first population) as there is no other sample population to compare it to. Agreed?

    This is an example of a "best fit" approach.

    The next population referenced is "Pima". The Pima people come from regions in Arizona, the opposite end of what is now the USA: http://en.wikipedia.org/wiki/Pima_people

    Again what the gedmatch calculator has done is found the next closest "best fit" population, the closest match based on the sample populations provided.

    This does not indicate that the "Clovis/Anzick" sample descends from the Pima people, but it could mean the opposite. The Pima people could descend from the Clovis people who occupied the area where the Anzick sample was discovered.

    You will note that the 4 population approximation has "@ 2.099648" at the end of it. This indicates that the "best fit" population averages are reasonably close to the Clovis sample.

    Once again these calculators are a best fit, and while they work for some people they will not provide the same results for others.

    To reiterate from my previous posts:

    Using Lochsbour as a sample population and comparing it to modern day Lithuanians will tell you far more than trying to do it the other way around.

    La Brana was not Irish, as much as you may want it to be, and I recall in the 4 population approximation was the result something in the order of "@ 14 or 16" or something like that?

    Using the La Brana sample as a sample population and comparing modern day Irish samples to it will tell you more than trying to compare in the reverse. There may be Irish people who descend from the same peoples as the La Brana sample. Or the "Irish average" could again simply be the best fit.

    I think Hinxton-4 is an interesting sample, and it would be worth having Hinxton-4 as a sample population and comparing modern Irish samples to it. Hinxton-4 may be an Iron Age Briton and there may be modern Irish people that descend from the people Hinxton-4 belonged to. The "Irish average" sample population may just be one of the "best fits" at this point for the Hinxton-4 sample, or the Hinxton-4 could be originally from Ireland...
    Last edited by N21163; 21 April 2015, 04:39 AM.

    Leave a comment:

Working...
X