Y-DNA-HAPLOGROUP-G-L Archives

From: <>
Subject: [Y-DNA-HAPLOGROUP-G] [Y-DNA-Haplogroup-G] G vs. G2 result:A possible reverse mutation of P15?
Date: Fri, 13 Oct 2006 22:09:38 -0700

With regard to the 2 postings by Carl and Lawrence today:

While a good variety of G subgroups as determined by SNPs would be the ultimate answer to determining origins, groupings and general migration patterns, those who are DYS388=12 seem to evolved into enough distinctive subgroups that the STR marker value combinations may prove just as valuable DYS388=13 persons, in contrast, overlap in their STR values so much that SNPs will be the only way to sort them out. Most of the 13s in Europe seem to have had a common male ancestor within 2500 years.

Unless many hundreds of G SNPs are found, it would seem that the STR marker values will still be needed to approximate relationships within, say, 2000 yrs within the same SNP group.

Ray Banks

---- Carl Feinstein <> wrote:
>
> Dear Ray,
>
> Once again I appreciate your thoughtful and logical approach and appreciate
> your information. It seems to me that you are listening to the data first,
> and only then checking historical time lines, to see if the data has
> congruence with historical events.
>
> In some of the other cases, it seems to me that pet historical hypotheses
> are coming first, and then the genetic data is being massaged to see if it
> can be made to confirm to these preferred hypotheses.
>
> However, more important than that, I am really scratching my head by what
> strikes me as insufficient attention to what molecular biology can tell us
> about the nature of nucleotide repeat sequences. The number of repeats in
> these sequences have widely varying stabilities from sequence to sequence.
> Some are very stable over many generations, while others undergo multiple
> mutations within the gametes or somatic cells of a single person over a few
> years. Furthermore, as the length or the sequences changes, the stability
> of the new sequences may not be the same as their predecessors. Therefore,
> to make these general assumptions about the rate of mutations for all the
> sequences seems to me to be without foundation, and probably not a useful
> approach.
>
> Also, I am increasingly dumfounded as to why people are still so
> over-focused on clustering larger and large numbers of repeat sequences.
> This approach is just a holding pattern of a technique until more SNP's of
> greater stability are tracked down. I think in 10 or 15 years people will
> look at these preoccupations with hunches about the mutation rates of
> clustered repeat sequences as a primitive, obsolete technique, useful only
> for a few transitional years before better techniques were developed. The
> further back we go utilizing these blind guesses and over-generalizations
> about mutation rates, the more inaccurate and speculative the calculations
> become.
>
> Regards,
> Carl Feinstein
>
>
>
>
>
>
>
> > This posting reply is actually not in reference to the subject in the title of
> > Ted's own posting, but rather in reference to the mentioned time sequences and
> > the relationships of the Jewish clades.
> >
> > TIME RELATIONSHIPS
> >
> > A few weeks ago was the first time I tried to make use of the time calculator
> > provided by Fluxus software. Ted has been very helpful in helping me to
> > understand how to use this portion of the software. At present, however, we
> > have not come to a concensus on the critical mutation rate to use for the
> > calculations.
> >
> > I have posted time calculation examples to a working page
> > http://members.cox.net/banksfamilies/TimeRelations.html
> >
> > As it happens, a few days ago I replaced Cano as an example and instead am now
> > using Ted as an example.
> >
> > I have a used a very conservative 250 yrs per mutation in a 37-marker sample.
> > The reason is to include the longest reach of the first standard deviation of
> > the time calculation and to put the Welsh clade in a logical timeframe. Ted,
> > in contrast, has been using about 120 yrs/mutation and posted diragrams using
> > the 67 marker sets, of which there currently are not many available. His
> > 67-marker networks do not include representatives from the secondary Jewish G
> > clade (Feinstein) and the possible crypto Jewish clade (Niccolo, Loccisano,
> > Cano) and the Sephardic Orol-Martin group unless I have overlooked these.
> >
> > You will note on my working page that the calculated distances from Ted to the
> > most recent common ancestor shared with these other groups are much longer
> > than the approximate 2,000 years Ted is advocating in his posting today. I
> > just recalculated these time relationships using Ted's approximate 118
> > yrs/mutation. This results in
> > Ted to Feinstein 6,959 yrs
> > Ted to Loccisano 4,366 yrs
> > Ted to Cano 3,717 yrs
> > Ted to Orol 4,366 yrs
> > ....all of these are well beyond the approximate 2,000 yrs Ted is advocating.
> >
> > I am having great difficulty with the usefulness of this software's time
> > calcuations. Several days ago I noticed that the addition of one more sample
> > to each of my 37-marker databases causes the time calculations to jump around
> > considerably except for tight clades whose common ancestor did not live more
> > than 2000-2500 yrs ago.
> >
> > There are also a few persons who have undergone some dramatic mutations at one
> > or the other marker, such a mutation seemingly too extreme to be explained by
> > the usual mutation process. There is a question in my mind whether such
> > outliers should be omitted from the database when doing time calculations.
> > They are likely to have some significant influence on the other listed time
> > calculations.
> >
> > The Welsh clade seems to be an unsually discrete group that has to be
> > accounted for in setting the mutation rate. Using my 250 yrs/mutation, they
> > seem to have a most recent common male ancestor about the time of the Norman
> > conquest (abt 1100 C. E.) Using Ted's faster rate, the common ancestor lived
> > in the 1400s-1500s. Our limited data show over 90% of available Welsh G men
> > of varying surnames belong to this one clade. One Jones from this group was
> > in Virginia in the 1600s, and most of the others with info show an ancestor
> > with their surname living in the late 1700s or early 1800s. Ted has pointed
> > out that some Welshmen were resisting taking permanent surnames well into the
> > 1800s, but the number is not quantified. This practice of lack of permanent
> > surnames does not seem to be the situation for those Welshmen who emigrated to
> > the U.S. and were using consistent surnames since the 1600s. Based on a
> > conservative estimate of haplogroup G in the Welsh population and consi!
> > dering that about 50% of Welsh G is showing up in America-only families, I
> > have estimated that this common ancestor for this Welsh G clade and his near
> > relatives living in the 1400s and 1500s in various parts of Wales would have
> > perhaps 100,000 G descendants today. This seems much more than would be
> > expected for 15-20 generations but not impossible for a small ancestral group
> > entering Wales in the middle Middle Ages.
> >
> > RELATIONSHIPS, JEWISH AND CRYPTOJEWISH CLADES
> >
> > I am a little surprised that Ted is still citing my chart from the summer
> > based on 21 markers that shows the large,(a) mostly Ashkenazi clade, (b) a
> > secondary Ashkenazi clade (Feinstein) and (c) a possibly cryto-Jewish grouping
> > (Loccisano, Cano) -- all three of them -- appearing together in one section of
> > the network chart. Several times I have privately made available the network
> > information from the 37-marker tests which now number over 250 samples. [ I
> > hope to eventually replace the 21-marker network charts with the 37-marker
> > ones if we can get a few more countries represented. The 37-marker charts have
> > 16 additional markers and would necessarily better show relationships.]
> >
> > In this 37-marker network, the Feinstein group has moved to another portion of
> > the chart and separated by other groups -- which is consistent with the
> > Feinstein clade now having the greatest time distance listed from Ted.
> > Because of the absence of some groups from these 37-marker samples, additional
> > moving around is to be expected when they are available, but I would be
> > surprised to see the Feinstein clade near to Ted ever again.
> >
> > In conclusion, my tentative conclusion is that the various Jewish &
> > cryptoJewish G groupings share most recent common ancestors who were living
> > over 4,000 yrs ago, perhaps much earlier, rather than about 2,000 yrs ago.
> >
> > Ellen Levy has been holding open the idea of these various Jewish G clades
> > originating in northeastern Europe rather than the Middle East soon before the
> > Current Era. Ted's large clade has so far failed to group with any of the
> > non-Jewish samples we have from that region. This type of n.e. European
> > origins [the Khazars?] cannot be ruled out for the Feinstein and Orol clades.
> > But it seems unlikely for the Loccisano-Cano group with an obvious more recent
> > Mediterranean origin. [Feinstein's time relationship to the large German G
> > clade (Bopp) is 3,127 yrs using Ted's fast mutation rate, but longer using my
> > rate under either mutation rate is half the time distance to Ted's clade.]
> >
> > Ray Banks
> >
> >
> > -------------------------------
> > To unsubscribe from the list, please send an email to
> > with the word 'unsubscribe' without
> > the quotes in the subject and the body of the message
>
>
> -------------------------------
> To unsubscribe from the list, please send an email to with the word 'unsubscribe' without the quotes in the subject and the body of the message

This thread:

• Re: [Y-DNA-HAPLOGROUP-G] [Y-DNA-Haplogroup-G] G vs. G2result:A possible reverse mutation of P15? by <>