GENEALOGY-DNA-L Archives

Archiver > GENEALOGY-DNA > 2010-06 > 1277393171


From: Jacques Beaugrand <>
Subject: Re: [DNA] mtDNA: HVR1, HVR2
Date: Thu, 24 Jun 2010 11:26:11 -0400
References: <4C224D2D.6103.5E9710E5@ng10066504jan.yahoo.co.uk>,<014601cb1308$cecdefe0$6c69cfa0$@net>,<BAY133-W25024890270B49C863C47FF1C60@phx.gbl><4C23458F.13997.626150B1@ng10066504jan.yahoo.co.uk>
In-Reply-To: <4C23458F.13997.626150B1@ng10066504jan.yahoo.co.uk>


Hello
In 2006 I have developed for the French Heritage a system to compare mtDNA
signatures
It can be seen à l'oeuvre at http://tinyurl.com/Frenchmtdna
Please be patient with these pages, the database comprises >1500 records,
it is a small home server and the bandwidth is rather limited.
You select the haplogroup by clicking on the letter corresponding to it
(left column).
I suggest you to select le letter H for which the French Heritage has 265
signatures.
These 265 will be listed in alphabetical order of the surname of their
owner.
Choose for instance Boivin kitno 95456 and click on the blue icon under
header Kin (column).
This will calculate an index of similarity with others in the database and
show them in order of relatedness.
In the present case 19 of them will show up possessing the same variations
in HVR1 and HVR2, with a K score of 1.00
(the more they share the same variations the more K is high; it is a % of
variations in common)
The signature you click moves to the top of the table and other signatures
are placed below it
in order of their *relatedness* with it, and a K index of 1.00 is shown
when they match perfectly.
Comparisons are made within major haplogroup, the H are compared to other
H, the A with the A, &c.
You can further click on the blue icon corresponding to another member in
the table.
For instance select 31300 Beaugrand, which are my mtDNA data.
Beaugrand happens to be H7 and will be compared only to other H7 in the
database
This program works by comparing signatures belonging to the same clade and
downstream when (sub-)subclades are present in the database.

The problems I encountered in comparing mtDNA variations of members in this
way is that
1. not all of them had their HVR1 or HVR2 typed, some had both, others only
HVR1.
2. the notation was implicit (519C rather than 16519C in the HVR1 region)
For ethical reasons we cannot use their FGS results when it is available and
this I deplore.
There could exist at least and anonymous database of FGSs with some
genealogy
beginning 2-3 generations upstream of the testee.
My code is cryptic and inelegant; but it still works.
I work with the Foxpro databases.
As concerns the following paragraph
> (I realize that not everyone is lucky enough to have just one sort of
> mtDNA in their bodies. One sequence of letters in every strand of
> their mtDNA.... but I'm working out the consequences of that in a
> separate discussion!)
it denotes a fine sense of humour of your part.
Salutations
[1]
On jeudi 24 juin 2010 at 06:46:23, you wrote :
> As you may know from reading other threads I am on a quest to work up
> a database to help people trying to find cousins.
> As I understand mtDNA and its testing we are already able to find the
> complete neucleotide ("letters") sequence for a given strand of
> mtDNA.
> At FTDNA, at least, customers can choose to have the "letters" in
> their mtDNA checked....
> EITHER....
> a)) ... just across two regions, and I quote from the FTDNA site:
> "HVR1(16001-16569) and HVR2(00001-00574)",
> ... OR ...
> b))... they can have the whole strand tested.
> (I realize that not everyone is lucky enough to have just one sort of
> mtDNA in their bodies. One sequence of letters in every strand of
> their mtDNA.... but I'm working out the consequences of that in a
> separate discussion!)
> ===
> My questions for this thread:
> Are those regions... HVR1 and HVR2... and their definitions broadly
> used, broadly agreed between most mtDNA researchers? Are there other
> labs out there offering to test the same ranges on the mtDNA? (I
> don't care whether the other labs, like FTDNA offer only "both
> regions or nothing", or whether they offer to do just one or the
> other region.)
> Thanks for what help you can offer.
> Tom
> Â
> -------------------------------
> 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

References

1. mailto:

This thread: