Page 95 of 441 FirstFirst ... 45859394959697105145195 ... LastLast
Results 941 to 950 of 4410

Thread: Waves of migration into South Asia

  1. #941
    Banned
    Posts
    5,171
    Sex

    A word about AG3. I'm sure you and others have noticed that it is called AG3_damage. Normally, when non-UDG treated aDNA is processed, we identify the number of bases from the 3' and 5' ends where we see evidence of post-mortem deamination in the form of C-->T and G-->A substitutions, and we trim a certain number of bases from the ends of reads to discard deaminated bases.

    With AG3, and with a couple of genomes I processed, the OPPOSITE was done, in other words discard the non-damaged bases, and retain the damaged bases.. It may sound crazy, but this is really the only option we have when we see that modern human DNA contamination is substantial, and here is why.

    Since we can't differentiate modern human DNA from aDNA, pretty much the only option we have left is to retain the damaged bases because we know those are ancient, and discard the non-damaged bases. This comes at a very high price though, because we can end up discarding 90% of the sequences. Not only that but we introduce some biases because we are only retaining T and A bases. This is what was done to AG3, hence the word damage. This is also what I have done with EHG_I0061_PMDS. In other words my definition of PMDS is equivalent to Fu et al's damage.

  2. The Following 3 Users Say Thank You to Kurd For This Useful Post:

     khanabadoshi (11-23-2017),  soulblighter (11-25-2017),  Varun R (11-24-2017)

  3. #942
    Banned
    Posts
    5,171
    Sex

    Sorted F3s in outgroup mode using Mbuti, with highest signal of admixture on top. Ukranian and Belarousan to simulate Russian modern human contamination. Malta is unremarkable and has expected ranking indicative of non-significant modern human DNA contamination. AG2 stands out as consistently more admixed than AG3, which is expected based on the author's estimates of 30-35% contamination. Margin of error for AG3 may be significant based on the authors retaining damaged fragments only

    F3 ( European; aDNA, Mbuti)

    SOURCE1 TARGET F3 STD ERR Z SNPs
    EHG_I0061_PMDS_DP2 Belarusian -0.0086 0.0020 -4.23 132940
    EHG_I0061_PMDS_DP2 Ukrainian -0.0077 0.0027 -2.80 129889
    EHG_I0061_PMDS_DP2 French -0.0027 0.0020 -1.35 134169
    EHG_I0061_ATLAS_DP2 Ukrainian 0.0023 0.0019 1.25 336030
    EHG Ukrainian 0.0023 0.0019 1.25 418294
    EHG_I0061_ATLAS_DP2 Belarusian 0.0037 0.0014 2.77 347646
    EHG Belarusian 0.0040 0.0014 2.95 435032
    EHG_I0061_ATLAS_DP2 French 0.0081 0.0013 6.34 352561
    EHG French 0.0087 0.0013 6.67 442649
    AfontovaGora2_Fu Belarusian 0.0104 0.0025 4.16 37988
    AfontovaGora2_Fu Ukrainian 0.0107 0.0032 3.35 36391
    AfontovaGora2_Fu French 0.0153 0.0024 6.45 38613
    EHG_I0061_PMDS_DP2 BedouinB 0.0167 0.0020 8.32 134252
    AfontovaGora3_Fu Belarusian 0.0167 0.0019 8.88 126771
    AfontovaGora3_Fu Ukrainian 0.0171 0.0025 6.73 121824
    Malta1_Fu Ukrainian 0.0175 0.0022 7.95 321793
    Malta1_Fu Belarusian 0.0183 0.0015 11.94 334888
    AfontovaGora3_Fu French 0.0207 0.0017 12.17 128950
    Malta1_Fu French 0.0222 0.0015 15.18 340808
    AfontovaGora2_Fu BedouinB 0.0246 0.0025 9.91 38880
    EHG BedouinB 0.0263 0.0013 19.96 444878
    EHG_I0061_ATLAS_DP2 BedouinB 0.0275 0.0013 20.85 352817
    Malta1_Fu BedouinB 0.0348 0.0015 23.57 342337
    AfontovaGora3_Fu BedouinB 0.0371 0.0018 20.85 129693

  4. The Following 2 Users Say Thank You to Kurd For This Useful Post:

     bmoney (11-24-2017),  Varun R (11-24-2017)

  5. #943
    Hi, Kurd,

    Interesting points about the Mal'ta Buret culture samples possibly being contaminated. I know that AG3 has the oldest example of the rs12821256 blonde allele. Is it possible that this could be from the contamination?

    Also, and I think this may be related to Kale's point about Basal Eurasian, but I've noticed that ADMIXTURE analyses don't show any of the neolithic farmer component in AG or Mal'ta. Doesn't that undermine the idea that it was contaminated significantly? Or am I missing something here?
    Last edited by arigiery; 11-24-2017 at 12:20 AM.

  6. #944
    Banned
    Posts
    5,171
    Sex

    Quote Originally Posted by arigiery View Post
    Hi, Kurd,

    Interesting points about the Mal'ta Buret culture samples possibly being contaminated. I know that AG3 has the oldest example of the rs12821256 blonde allele. Is it possible that this could be from the contamination?

    Also, and I think this may be related to Kale's point about Basal Eurasian, but I've noticed that ADMIXTURE analyses don't show any of the neolithic farmer component in AG or Mal'ta. Doesn't that undermine the idea that it was contaminated significantly? Or am I missing something here?
    Yes, quite possible. First, the sample yielded only about 1% endogenous DNA, and the average coverage was only 0.17, and it is a pseudo-haploid, meaning that 1 of 2 alleles is wrong at each hetrozygous position. Add to that only damaged segments were retained because of estimated significant modern human DNA contamination.

    The Gedmatch calculators are the worst tools I can think of for this type of purpose. Not only that, since the basal admixed aDNA we have is all younger than the 16.5K year old AG3, how can basal geneflow to AG3 be determined using ADMIXTURE even if she had any.


    EDIT: I just combed through AG3's genome and didn't see that she was genotyped at rs12821256. Where did you get that info?



    EDIT 2: GEDMATCH IDs: AG3- Z303443- 234K SNPs
    MA1- Z318302- 696K SNPs
    Last edited by Kurd; 11-24-2017 at 01:22 AM.

  7. The Following User Says Thank You to Kurd For This Useful Post:

     bmoney (11-24-2017)

  8. #945
    Quote Originally Posted by Kurd View Post
    Yes, quite possible. First, the sample yielded only about 1% endogenous DNA, and the average coverage was only 0.17, and it is a pseudo-haploid, meaning that 1 of 2 alleles is wrong at each hetrozygous position. Add to that only damaged segments were retained because of estimated significant modern human DNA contamination.

    The Gedmatch calculators are the worst tools I can think of for this type of purpose. Not only that, since the basal admixed aDNA we have is all younger than the 16.5K year old AG3, how can basal geneflow to AG3 be determined using ADMIXTURE even if she had any.


    EDIT: I just combed through AG3's genome and didn't see that she was genotyped at rs12821256. Where did you get that info?



    EDIT 2: GEDMATCH IDs: AG3- Z303443- 234K SNPs
    MA1- Z318302- 696K SNPs
    Info on AG3's blonde gene is on page 50-51 of the Mathieson 2017 paper...which in turn references Fu et al's "Genetic History of Ice Age Europe" Paper. I don't have access to the fu paper at the moment.

    Forget about "basal Eurasian" for a minute, since that's a hypothetical ghost population....we have no ancient pure sample of "Basal Eurasian." But we do have samples of neolithic Anatolia that we are pretty sure contributed to all modern European gene pools (and this clearly shows up in ADMIXTURE). What I was saying is that, if AG3 (or AG2, or Mal'ta boy) were contaminated with modern Russian DNA, I would expect to see the "Anatolian neolithic farmer" component that all modern Europeans (including Russians) have in the ADMIXTURE. Yet I don't. What I *do* see is the "Eastern Hunter Gatherer" component, which I suspect is drawing it towards modern European samples.

    In short, what I'm asking is this: if the Mal'ta Buret culture samples have modern European contamination, why does ADMIXTURE show them missing a key component of the modern european gene pool?

  9. #946
    Quote Originally Posted by Kurd View Post
    Sorted F3s in outgroup mode using Mbuti, with highest signal of admixture on top. Ukranian and Belarousan to simulate Russian modern human contamination. Malta is unremarkable and has expected ranking indicative of non-significant modern human DNA contamination. AG2 stands out as consistently more admixed than AG3, which is expected based on the author's estimates of 30-35% contamination. Margin of error for AG3 may be significant based on the authors retaining damaged fragments only

    F3 ( European; aDNA, Mbuti)

    SOURCE1 TARGET F3 STD ERR Z SNPs
    EHG_I0061_PMDS_DP2 Belarusian -0.0086 0.0020 -4.23 132940
    EHG_I0061_PMDS_DP2 Ukrainian -0.0077 0.0027 -2.80 129889
    EHG_I0061_PMDS_DP2 French -0.0027 0.0020 -1.35 134169
    EHG_I0061_ATLAS_DP2 Ukrainian 0.0023 0.0019 1.25 336030
    EHG Ukrainian 0.0023 0.0019 1.25 418294
    EHG_I0061_ATLAS_DP2 Belarusian 0.0037 0.0014 2.77 347646
    EHG Belarusian 0.0040 0.0014 2.95 435032
    EHG_I0061_ATLAS_DP2 French 0.0081 0.0013 6.34 352561
    EHG French 0.0087 0.0013 6.67 442649
    AfontovaGora2_Fu Belarusian 0.0104 0.0025 4.16 37988
    AfontovaGora2_Fu Ukrainian 0.0107 0.0032 3.35 36391
    AfontovaGora2_Fu French 0.0153 0.0024 6.45 38613
    EHG_I0061_PMDS_DP2 BedouinB 0.0167 0.0020 8.32 134252
    AfontovaGora3_Fu Belarusian 0.0167 0.0019 8.88 126771
    AfontovaGora3_Fu Ukrainian 0.0171 0.0025 6.73 121824
    Malta1_Fu Ukrainian 0.0175 0.0022 7.95 321793
    Malta1_Fu Belarusian 0.0183 0.0015 11.94 334888
    AfontovaGora3_Fu French 0.0207 0.0017 12.17 128950
    Malta1_Fu French 0.0222 0.0015 15.18 340808
    AfontovaGora2_Fu BedouinB 0.0246 0.0025 9.91 38880
    EHG BedouinB 0.0263 0.0013 19.96 444878
    EHG_I0061_ATLAS_DP2 BedouinB 0.0275 0.0013 20.85 352817
    Malta1_Fu BedouinB 0.0348 0.0015 23.57 342337
    AfontovaGora3_Fu BedouinB 0.0371 0.0018 20.85 129693
    All interesting, but how does this necessarily show recent admixture in Af2 and maybe to a lesser extent AF3? Couldn't it also be that Af2 shares more drift with Slavs than the other two do? Or Af2 contributed more to modern Slavs' ancestry?

  10. #947
    Banned
    Posts
    5,171
    Sex

    Quote Originally Posted by arigiery View Post
    All interesting, but how does this necessarily show recent admixture in Af2 and maybe to a lesser extent AF3? Couldn't it also be that Af2 shares more drift with Slavs than the other two do? Or Af2 contributed more to modern Slavs' ancestry?
    I would expect to see the "Anatolian neolithic farmer" component that all modern Europeans (including Russians) have in the ADMIXTURE. Yet I don't. What I *do* see is the "Eastern Hunter Gatherer" component, which I suspect is drawing it towards modern European samples.

    You should have read my posts more carefully, because I just wrote in post 940 :
    Dstats are very useful but clearly not for this purpose, because there is so many things besides basal Eurasian admixture that can affect the outcome. First, the MA1 split is closer to the Ust split than splits such as WHG are to Ust. Second, it is impossible to assess how "drifted" the various WHGs are in comparison to MA1, and one can go on. Each of these factors can act to increase or decrease shared alleles with Ust.
    and suggested there if you really want to use stats:
    If you want to use dstats, then it is best to compare members of the same clade (MA1, AG2, AG3) against various genomes to assess basaleness, Russian shift, W European shift, etc. I ran a few for you which you can archive for future use. You may be able to discern patterns although it is still impossible to quantify the amount of modern DNA contamination based on these, but you can certainly pick up oddities. To gauge basalness, I used the imputed Satsurbila, the high coverage (10X) Iran WC1, and Natufian. To assess Russian shift, I used Belarousans, and to assess W European shift I used French.
    even then you can not quantify amounts and will miss smaller amounts altogether.


    Scientists use the Mt DNA and the X chromosome (in males only) to quantify the amount of modern human contamination, and I have use the X chromosome in all the male steppe samples I personally diploid genotyped which I have annotated with PMDS. If you read Raghavan 2014, Upper Palaeolithic Siberian genome reveals dual ancestry of Native Americans supp, you will see that they estimate AG2's modern contamination using these techniques at 30-35%. which you'll get an indication for by looking at the f3s and dstats I posted (although stats can't be used to estimate amounts). You'll see that AG2 shares signifiantly more drift with Bellarousans and Ukranians (Russian proxies) and basal admixed aDNA than AG3.

    A little more detail on the techniques we use to estimate modern human contamination in aDNA:

    X CHROMOSOME (MALES ONLY):

    Since males have 1 X and are expected to be haploid on their X, you count the number of positions where the aDNA is diploid, and use that to estimate contamination in the nuclear DNA. If you check my website you'll see that I used the DICE program introduced by Renaud et al to estimate modern human contamination in the steppe samples I diploid genotyped. Then I used a technique introduced by Skoglund, PMDS to correct for contamination (those steppe samples are marked PMDS). This is essentially what Fu et al did with AG3 hence it is annotated as AG3_damage.

    As crazy as it may seem, discarding undamaged segments in aDNA, and retaining damaged ones is pretty much the only option we have for samples where we estimate high modern human contamination. I personally only use it as a last resort because it discards as much as 90+% of the sequences, and IMO introduces a bias, but that is what Fu et al did in almost 1/2 of their published ancient samples (check their supp). The margin of error increases when we do this (AG3), but it also increases due to sequencing errors introduced by things such as low coverage ( mean coverage AG2-only 0.071, AG3- only 0.17, MA1- 1.17. None of these are high enough for me to diploid genotype), and also because 90%+ of the published aDNA is pseudo-haploid (1 wrong allele at every hetrozygous) position. This is why I diploid genotype whenever mean coverage allows me to.


    mtDNA;

    Fu et al used this to estimate modern human contamination in their female samples as well as male sample along with the X chromo technique (males only). They used this to declare a sample contaminated when:

    1- The fraction of fragments matching the reconstructed consensus better than any of 311 worldwide mtDNA sequences used for comparison is <95%. OR
    2- The 95% confident lower bound of the fraction of fragments matching the consensus better than any of 311 worldwide mtDNA sequences is <85%.

    In short, what I'm asking is this: if the Mal'ta Buret culture samples have modern European contamination, why does ADMIXTURE show them missing a key component of the modern european gene pool?
    I can spend 3 pages explaining this to you, and I don't have time. Best to read the material I have published on my website
    Last edited by Kurd; 11-24-2017 at 01:01 PM. Reason: Typo-changed Renaud to Skoglund

  11. The Following 2 Users Say Thank You to Kurd For This Useful Post:

     bmoney (11-26-2017),  palamede (11-25-2017)

  12. #948
    Registered Users
    Posts
    1,742
    Sex
    Omitted

    Quote Originally Posted by Kurd View Post
    Dstats are very useful but clearly not for this purpose, because there is so many things besides basal Eurasian admixture that can affect the outcome. First, the MA1 split is closer to the Ust split than splits such as WHG are to Ust. Second, it is impossible to assess how "drifted" the various WHGs are in comparison to MA1, and one can go on. Each of these factors can act to increase or decrease shared alleles with Ust.
    Drift is not an issue...Ust-Ishim doesn't care if you are a Native American from now or from 12,000 years ago.
    Chimp Ust_Ishim Karitiana Clovis 0.0024 0.380 352350

  13. #949
    Quote Originally Posted by Kurd View Post





    You should have read my posts more carefully, because I just wrote in post 940 :


    and suggested there if you really want to use stats:


    even then you can not quantify amounts and will miss smaller amounts altogether.


    Scientists use the Mt DNA and the X chromosome (in males only) to quantify the amount of modern human contamination, and I have use the X chromosome in all the male steppe samples I personally diploid genotyped which I have annotated with PMDS. If you read Raghavan 2014, Upper Palaeolithic Siberian genome reveals dual ancestry of Native Americans supp, you will see that they estimate AG2's modern contamination using these techniques at 30-35%. which you'll get an indication for by looking at the f3s and dstats I posted (although stats can't be used to estimate amounts). You'll see that AG2 shares signifiantly more drift with Bellarousans and Ukranians (Russian proxies) and basal admixed aDNA than AG3.

    A little more detail on the techniques we use to estimate modern human contamination in aDNA:

    X CHROMOSOME (MALES ONLY):

    Since males have 1 X and are expected to be haploid on their X, you count the number of positions where the aDNA is diploid, and use that to estimate contamination in the nuclear DNA. If you check my website you'll see that I used the DICE program introduced by Renaud et al to estimate modern human contamination in the steppe samples I diploid genotyped. Then I used a technique introduced by Skoglund, PMDS to correct for contamination (those steppe samples are marked PMDS). This is essentially what Fu et al did with AG3 hence it is annotated as AG3_damage.

    As crazy as it may seem, discarding undamaged segments in aDNA, and retaining damaged ones is pretty much the only option we have for samples where we estimate high modern human contamination. I personally only use it as a last resort because it discards as much as 90+% of the sequences, and IMO introduces a bias, but that is what Fu et al did in almost 1/2 of their published ancient samples (check their supp). The margin of error increases when we do this (AG3), but it also increases due to sequencing errors introduced by things such as low coverage ( mean coverage AG2-only 0.071, AG3- only 0.17, MA1- 1.17. None of these are high enough for me to diploid genotype), and also because 90%+ of the published aDNA is pseudo-haploid (1 wrong allele at every hetrozygous) position. This is why I diploid genotype whenever mean coverage allows me to.


    mtDNA;

    Fu et al used this to estimate modern human contamination in their female samples as well as male sample along with the X chromo technique (males only). They used this to declare a sample contaminated when:

    1- The fraction of fragments matching the reconstructed consensus better than any of 311 worldwide mtDNA sequences used for comparison is <95%. OR
    2- The 95% confident lower bound of the fraction of fragments matching the consensus better than any of 311 worldwide mtDNA sequences is <85%.



    I can spend 3 pages explaining this to you, and I don't have time. Best to read the material I have published on my website
    I revisited your post where you tried to detect basal-ness using Natufian and Iran Neolithic. This does make me a bit suspicious of AG2 at least. BTW I agree with your point that you made a while back, where you said that it's probably best to assume that analyses of our three Mal'ta Buret samples are possibly leading us to overestimate the European affinity of ancient Siberians. "True ANE" *might* be something more Amerindian and Kalash like.

    Come to think of it, I haven't seen AG2 or AG3 discussed/used as much as Mal'ta, and it sounds like there is good reason for this.

  14. The Following 2 Users Say Thank You to arigiery For This Useful Post:

     bmoney (11-26-2017),  Kurd (11-24-2017)

  15. #950
    Banned
    Posts
    5,171
    Sex

    Quote Originally Posted by Kale View Post
    Drift is not an issue...Ust-Ishim doesn't care if you are a Native American from now or from 12,000 years ago.
    Chimp Ust_Ishim Karitiana Clovis 0.0024 0.380 352350
    Seriously...how do you think Eurasians went from 1 or a few out-of-Africa cluster to 100s of clusters you see now.You may want to familiarize yourself with the concept of genetic drift and how it is by far a more important force than random mutations. I have an article on my website where I have discussed this at http://www.eurasiandna.com/2017/11/0...adoxical-onge/

    Mutations can act to increase diversity, but their rate is only about 10-8 /bp/generation for humans, which is sort of slow. For example, with 3 x 109 base pairs in the genome, this translates to about 30 mutations per genome per generation, or alternatively, 0.01 mutations per 1,000,000 SNPs ( typical public datasets) per generation, or alternatively, only 16 mutations per genome per 40,000 years (assuming 25 years per generation). Clearly, mutations are a weak force and alone can’t counteract the rapid rate at which alleles either become lost or fixed in small drifted populations such as the Onge.

    Assuming no selection pressures on an allele, which is usually the case with non-functional intergenic alleles, and genetic drift is the only evolutionary force acting on it, the probability that the allele will eventually become fixed is simply its allele frequency in the population. So for example if allele A has a frequency of 85% within the Onge, then the probability that A will become fixed in due time is 85%, and the probability that A will be lost in due time may be 15%. Generally, my analysis has shown that in small drifted populations, alleles with low frequencies tend to become lost, and alleles with high frequencies tend to become fixed in due time.

    According to the Wright–Fisher model, the following formula can be used for approximating the time it takes a neutral allele to become fixed via genetic drift:

    T fixed = [ -4Ne (1 – p) ln (1 – p)] / p

    where T is the time in generations, p the allele frequency, and Ne the effective breeding population size.



    EDIT: Although I am not the authority on dstats (Nick Patterson is since he was involved in the creation of the program), I believe the dstat you posted only shows that Karitiana share the same amount of drift as Clovis with Ust. In other words, the drift paths of Clovis and Karitiana individually, post split of the native American branch from the tree do not affect the dstat. You may want to check the stat using various Amerindians to rule out coincidence
    Last edited by Kurd; 11-24-2017 at 07:29 PM.

  16. The Following 3 Users Say Thank You to Kurd For This Useful Post:

     bmoney (11-26-2017),  parasar (11-24-2017),  soulblighter (11-25-2017)

Page 95 of 441 FirstFirst ... 45859394959697105145195 ... LastLast

Similar Threads

  1. Replies: 9
    Last Post: 05-21-2020, 09:38 AM
  2. Replies: 8
    Last Post: 11-16-2018, 09:16 PM
  3. Replies: 0
    Last Post: 02-17-2017, 04:01 AM
  4. Replies: 13
    Last Post: 02-26-2013, 08:43 PM

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •