Results 1 to 5 of 5

Thread: I1 SNP Rate

  1. #1
    Registered Users
    Posts
    495

    I1 SNP Rate

    If I1 has a formation date of 27,500 ybp and a TMRCA of 4,600 ybp and it has 306 SNPs that are unique, wouldn’t that give I1 an average SNP mutation rate of around 75 years per SNP? If that is the case why do YFULL and other scientists use mutation rates that are between 135-160 years per SNP? What am I misunderstanding here? I am aware that particular paternal lines can have more or less SNPs and that the variability can be 1-10 generations per SNP. I’m talking about overall averages.
    I1> I-DF29> I-Z58> I-Z59> I-Z2041> I-Z2040> I-Z382> I-S26361>
    I-S16414> I-FGC24354>
    I-FGC24357> I-S10350> I-Y125947>
    I-BY149414 TMRCA 1,400 ybp

    Scandinavia-> British Isles-> Belgium-> Canada-> United States

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

     C J Wyatt III (10-17-2018),  JMcB (10-18-2018),  JonikW (10-17-2018)

  3. #2
    Registered Users
    Posts
    392
    Sex
    Location
    United Kingdom
    Ethnicity
    European
    Nationality
    British
    Y-DNA
    I-L338
    mtDNA
    J1c1

    United Kingdom England England North of England Norfolk Scotland Ireland
    YFull's age estimation methodology is explained here: https://www.yfull.com/faq/what-yfull...n-methodology/ and here https://www.researchgate.net/publica...equencing_Data

    It's important to realize that not every SNP is counted for the age estimation calculation. Most of these need to be in the combBED region, so SNPs from outside that region are not counted for age estimation. There are other criteria as well - indels, SNPs that appear in many locations, SNPs that are homologous, SNPs that are low quality. Read the papers in the links that I posted above. If you have YFull analysis, you can check your own results under the Age Estimation tab and see which known and novel SNPs are accepted and excluded.

  4. The Following 3 Users Say Thank You to deadly77 For This Useful Post:

     Dave-V (10-18-2018),  JMcB (10-17-2018),  JonikW (10-18-2018)

  5. #3
    Registered Users
    Posts
    205
    Sex
    Location
    USA
    Nationality
    USA
    Y-DNA
    R1b-L21 L513*

    United States of America Ireland Germany Belgium Wallonia
    As an example of deadly77's explanation, if you go to YFull's parent tree for I1 at https://www.yfull.com/tree/IJ/, and click on the "info" box for I (the parent of I1), you'll see the calculation details for the 27,500 age estimate. In the second column, the number of SNPs used varies in the 150-180 range (not including outliers) and that's down to present day, not just the I1 block of 306 SNPs. Since not all discovered SNPs are used in age calculation, the mutation rate used in the formula is higher than the frequency of known SNPs.

    Another factor is that the 306 SNPs represents the total SNPs discovered in that section of the tree from a large sample size, and due to coverage and read quality variations no single test will discover all 306, so with individual tests you're already starting with a smaller number of SNPs.

  6. The Following 3 Users Say Thank You to Dave-V For This Useful Post:

     deadly77 (10-18-2018),  JMcB (10-18-2018),  JonikW (10-18-2018)

  7. #4
    Registered Users
    Posts
    495

    Interesting! Thanks for the replies! I have another question. I’m expecting my Big Y-500 SNP results any day now. When I eventually submit my data to YFULL will they recognize a similar amount of variants? Do FTDNA and YFULL use similar criteria when deciding what variants are recognized?

  8. The Following 3 Users Say Thank You to mwauthy For This Useful Post:

     Adrian Stevenson (10-19-2018),  JMcB (10-18-2018),  JonikW (10-18-2018)

  9. #5
    Registered Users
    Posts
    392
    Sex
    Location
    United Kingdom
    Ethnicity
    European
    Nationality
    British
    Y-DNA
    I-L338
    mtDNA
    J1c1

    United Kingdom England England North of England Norfolk Scotland Ireland
    Quote Originally Posted by mwauthy View Post
    Interesting! Thanks for the replies! I have another question. I’m expecting my Big Y-500 SNP results any day now. When I eventually submit my data to YFULL will they recognize a similar amount of variants? Do FTDNA and YFULL use similar criteria when deciding what variants are recognized?
    They will be similar, but not the same. The two companies do use some different criteria for SNPs that they call. The biggest differences appear to be in number of reads and position on the Y chromosome.

    With regard to the first one, this blog by Roberta Estes might be heplful https://dna-explained.com/2018/01/12...-results-hg38/ - it's mostly about the hg19 to hg38 conversion, but if you scroll down to about quarter of the way down the page to where it says "Confidence Levels" - main takeaway is that 10 reads of a SNP are high confidence calls, 5-8 reads of a SNP are medium confidence calls (9 reads not mentioned )and high and medium confidence calls are the ones that they'll report. It does say that they will use SNPs with a lower threshold for a number of reads in some circumstances such as if it's phylogenetically consistent with other samples or reported by projects admins or third party analysis. So it's not a hard and fast rule and there is some room to maneuver. As far as I'm aware, Roberta Estes isn't an employee of FTDNA, but she does work with them and she thanks Michael Sager (who is an employee of FTDNA) for assistance with the blog article, and she is usually very thorough in her blog, so I think this is a good representation of how FTDNA is actually calling SNPs (or at least how they were doing it in January). In contrast, it seems that YFull will call a lot more SNPs with lower number of reads. So I guess that you could say that YFull is less conservative than FTDNA in this regard.

    Another major difference is that YFull don't seem to be reporting new SNPs in certain regions of the Y chromosome such as in the centromere and the DYZ19 repeat section, while FTDNA seems to report quite a few SNPs in these regions. Again, not a hard and fast rule - I've seen examples of YFull including SNPs from the DYZ19 region when people have asked to include them because it's important for branching of the tree. So I guess in this regard you could say that FTDNA is less conservative than YFull.

    Regardless of how they are reported, you will be able to look at the SNPs in question using the BAM viewer at either site.

    Probably be easier to go through the differences once your results and analysis are done with both and then you can look at the differences and try to rationalize them.

  10. The Following 3 Users Say Thank You to deadly77 For This Useful Post:

     JMcB (10-18-2018),  JonikW (10-18-2018),  palamede (10-19-2018)

Similar Threads

  1. Average of SNP (and/or Indels) mutation rate
    By palamede in forum General
    Replies: 19
    Last Post: 10-14-2018, 09:43 PM
  2. How would you rate MyGenomeBox? (+ post your results)
    By KingKhalasi in forum Autosomal (auDNA)
    Replies: 41
    Last Post: 09-04-2018, 07:00 AM
  3. Genotype rate - Would you explain this term?
    By Lupus82 in forum Inquiries Corner
    Replies: 6
    Last Post: 01-31-2018, 11:47 AM

Posting Permissions

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