GENEALOGY-DNA-L Archives

Archiver > GENEALOGY-DNA > 2005-06 > 1119030758


From: Terry Todd <>
Subject: basing conclusions on a sample of 1
Date: Fri, 17 Jun 2005 12:52:38 -0500
References: <000001c572e8$a621a890$041e9951@default> <024a01c57362$e1d8acb0$1d78190a@scvwd.gov>
In-Reply-To: <024a01c57362$e1d8acb0$1d78190a@scvwd.gov>; from rmcmurtry@igc.org on Fri, Jun 17, 2005 at 10:34:58AM -0700


I am forwarding an email exchange from one of our group members
for this list to ponder on.

Thanks,
Terry Todd


On Fri, Jun 17, 2005 at 10:34:58AM -0700, Richard McMurtry wrote:
> Dear Alastair etal,
>
> Thanks so much for your reply! However, I think there was one point that
> was misleading about the attached chart that I need to clarify in order to
> ask questions 5 and 6 correctly.
>
> In the family of Andrew Todd d 1791, we know he had a son JOhn, but we don't
> know whether John 1750-1813 was that son. See "?" added to chart attached.
>
> So, without that certainty, we only have one sample that is confirmed to be
> a desc of Andrew Todd d 1791, namely, Jim C. Todd with a "13" in marker 461.
> The only family with male descendants until the present generation are the
> ones on the chart. So, our question becomes:
>
> Q5: Given that we have only one sample for a proven descendant of Andrew d
> 1791 namely Jim C and this Jim C had a 13 in marker 461, and if we were to
> get a sample for Scott B that showed a 13 in marker 461, what would be the
> relative probability that the mutation occured (1) in the generation of
> Robert Scott 1851-1919, (2) in the generation of Wiliam 1827-1877, (3) in
> the gerneration of Robert 1807-1830, (4) in the geneeration of Robert
> 1762--1829, (5) in the generation of Andrew d 1791? Do you see my dilemna?
> I can't tell whether getting a second sample so far down in the chart will
> do me any good at showing the probabilities that the mutation occured in
> Andrew's generation? That's what I need to show - that the mutation
> happened in Andrew's generation. If I can't show that the mutation occured
> in Andrew's generation, that I can't claim that John 1750-1813 is the son of
> Andrew.
>
> Is it as simple as saying that there are 5 individuals who could have had
> the mutation and it is equally likely that it occured in any of these
> generations and hence, the chance of the mutation having happend in
> Andrew's generation is only 1/5? In other words, not very likely.
>
> Q6: Similarly this would revise Question 6 to be: Given that we have only
> one sample for a proven descendant of Andrew d 1791 namely Jim C and this
> Jim C had a 13 in marker 461, and if we were to get a sample for Jim D. that
> showed a 13 in marker 461, what would be the relative probability that the
> mutation occured (1) in the generation of William 1827-1877, (2) in the
> gerneration of Robert 1807-1830, (3) in the geneeration of Robert
> 1762--1829, (4) in the generation of Andrew d 1791? Is it as simple as
> saying that there are 4 individuals who could have had the mutation and it
> is equally likely that it occured in any of these generations and hence, the
> chance of the mutation having happened in Andrew's generation is only 1/4?
> In other words, no more likely than any other generation.
>
> What is so ironic about this data is that we appear to have been very very
> lucky and the mutation that distinguishes the Andrew line happened exactly
> when we needed it to in order to prove that Andrew was a 13. But we can't
> prove this since Andrew's other sons (not shown on chart) had no male
> descendants living in the present day and Andrew's son Robert had only one
> son (as shown on the chart) that had male descendatns living in the present
> day. We seem to be so close to having proof, yet not close enough.
>
> Could you please take a shot at Questions 5 and 6 again in light on the
> clarification about John 1750-1813 not being proved to be the son of Andrew
> d 1791?
>
> Richard McMurtry
>
>
>
>
> ----- Original Message -----
> From: "DNA Heritage" <>
> To: <>; <>
> Sent: Thursday, June 16, 2005 7:59 PM
> Subject: Fw: Questions on data interpretation
>
>
> Hello Richard,
>
> As you will know, mutations are very slow. Even the 'fast' markers are
> incredibly slow. Expect a mutation anywhere between one in 250-500
> generations, which you will appreciate can and does happen. Therefore, the
> most important thing to do is not to place too much value on the precise
> stats.
>
> Q1
> If you have two brothers who are a value at 12, and one brother that has a
> value of 11, you can be very certain that the father was a 12 also. Imagine
> that the ancestor was an 11. For an identical mutation (i.e. an increase of
> one) to happen on the same marker within the same generation would be
> extremely slim - approx. 1 in 500,000 (using a 0.002 mutation rate per
> marker, a single-step mutation model and equal chance of an increase or
> decrease of one). Conversely, the probability that the ancestor was
> actually a 12 would be approaching 1.
>
> This should give you an idea of the likelihood of a change in the marker.
>
> Q2
> Approx 0.93
>
> Q3
> Similar to Q1. Approaching 1 that the father of Robert, Andrew and William
> is 12 at DYS461
>
> Q4
> If the tree is correct, the again, this is approaching 1 also.
>
> As I stated at the top of this letter, the mutations do happen very slowly
> and for them to happen twice on the same marker within a short time period
> would be extremely rare.
>
> This calculator will no doubt have you playing with figures:
> http://members.aol.com/dnafiler/MutationCalculator.exe
>
> The chance of observing a 13 in Scott is also high - so this shouldn't be
> the thrust of testing - determining if he is indeed part of the predicted
> line should be.
>
> For your last question, again, the probability is very high. I would agree
> with your assumption that the mutation did indeed happen when you believe it
> to have done.
>
>
>
>
> Kind Regards,
> Alastair
>
>
>
> Alastair Greenshields
> Principal
> DNA Heritage
> http://www.dnaheritage.com
>
> __________________________________
>
>
> Dear Alastair et al,
>
>
>
>
>
>
> I have been working with you for a couple of years it seems and now need
> some help interpreting the data.
>
>
>
> I am writing to ask for your help with specific questions relative to
> understanding my DNA results.
>
>
>
> Enclosed are:
>
> A Genealogical Chart with the results for Marker 461
> A spreadsheet showing the complete DNA marker data AND showing just the
> markers that differ
>
>
> Please note that Andrew and William did have sons named John as shown on the
> chart, but we don't know for sure whether John 1750-1813 or John 1746-1829
> is the son of Andrew and whether John Todd 1746-1829 is the son of William.
>
>
>
> My questions are:
>
> Given that a descendant of both Samuel and Low (sons of William) have a 461
> value of 12 , what is the relative probability that William had a 12 at 461?
>
> Given that a descendant of Samuel and Low (sons of William) have a 461 value
> of 12, what is the relative probability that William's son John also has a
> 12 at 461 compared to someother value at 461?
> Given that a descendant of Robert and a descendant of William had 12 at 461
> and a descendant of Andrew had a 13 at 461, what is the relative probability
> that the father of them all had a 12 at 461 compared to a 13?
> Given that a descendant of Andrew had a 13 at 461, what is the relative
> probability that the mutation happened in the generation of Andrew himself
> compared to each of the succeeding generations between Andrew and the sample
> donor?
> If we were to get a second sample from a descendant of Andrew in the
> generation labelled Scott on the chart, what change would this give us in
> the relative probability that Andrew d 1791 was a 13?
> If we were to get a second sample from a desc. of Andrew d 1791 in the
> generation labelled Scott on the chart, what would be the relative
> probability that Andrew's son John was a 13 compared to being another value?
>
> Given that a descendant of both sons of John 1750-1813 have 13 in marker
> 461, what is the relative probability that John had a 13 in marker 461 as
> compared to a 12?
>
>
>
>
> Note that there are no other male descendants of Andrew other than those in
> the generation indicated on the chart.
>
>
>



This thread: