Archiver > GENEALOGY-DNA > 2011-10 > 1319385265

From: Jørgen K. Kanters <>
Subject: Re: [DNA] How should ISOGG define "private SNP" and "provisionalSNP"?
Date: Sun, 23 Oct 2011 17:54:25 +0200
References: <A6329B9FD895498A876CB51FE563B27C@PC>
In-Reply-To: <A6329B9FD895498A876CB51FE563B27C@PC>

In principle I agrees with you. However I think it is a minor problem, and
the real documented private mutation probably do not exist since most people
will not test for a private mutation. Furthermore a confirmed private
mutation is only private until other subjects are found later on with the
same private mutation. And then how related should two subjects with a SNP
be, when still conceived as a private mutation? TMCA=100 y 200y or another?

As long as ISOGG have they own but precise terminology, It is IMHO not a


2011/10/22 Lancaster-Boon <>

> Hello everyone
> I have come across a subject which seems to need some community discussion.
> I am going to present my own opinions, but I'd be very open to hear
> well-reasoned opposed positions if there are any.
> In some discussions I was involved in recently I have learnt that at least
> one adminstrator of ISOGG's Y phylogeny webpages, uses a definition of
> "private SNP" which seems surprising, at least to me. And he claims this is
> a decided usage within ISOGG which is unlikely to be changed. He uses this
> definition when categorizing SNPs on the phylogeny webpages. These have
> become very important to the community of genetic genealogists.
> The discussion came about because of a concern raised by another person
> that
> the definition being used may well make it very difficult for members of
> ISOGG (genetic genealogists) to judge what tests to invest in. And indeed,
> it appears to most discourage the testing of the SNPs which should be most
> tested.
> Here are some quotes from the adminstrator involved:
> "I believe that you misunderstand what our definition of private. Private
> merely means that we do not yet have enough data to support the position
> that a particular SNP is found in significant numbers in a known population
> or have high enough STR diversity."
> "The ISOGG guidelines state a SNP is classified as private until enough
> data
> has been presented to the committee that satisfy our guidelines for an
> upgrade to 'confirmed within subclade', 'provisional' or 'fully
> confirmed'."
> "The provisional status is for what we see as tree worthy markers. It is
> not
> a category for markers observed once and/or being researched. For such
> markers, we have the private status. It is not possible for us to add
> eleven
> SNPs to the tree as provisional subclades without strong supporting data."
> *"I don't have time to constantly argue what private is or what is not. You
> are risking becoming redundant. I've stated and settled my personal
> position
> and the ISOGG's definition of it."
> I have two concerns, which I know are shared by others:-
> 1. The style of responses indicates a new attitude I had never thought I
> would see in ISOGG. There are committees who decide things outside of the
> community and we should not dare to question this. It seems like a
> disconnect between ISOGG and GG. But what is ISOGG for if it sees itself as
> somehow opposed to its genetic genealogist members?
> 2. The definition being used seems to be as follows:
> *Provisional means already pretty well studied, but not yet definitively
> cross checked, and probably a significant clade defining SNP.
> *Private means hardly studied at all, and therefore quite often this term
> is
> being used to describe SNPs which will be found to be very significant.
> Indeed ISOGG's webpages, also contains a definition which can be found
> here:-
> ===
> *Provisional BPs are color coded and defined as: BPs newly submitted to the
> ISOGG committee that have sufficient information to be placed in the tree
> with some certainty, but insufficient to meet all the criteria above and
> are
> awaiting evidence as described above.
> *Private SNPs are color coded and are defined as:
> --EITHER a BP that has been observed only once, or has been observed
> multiple times but the associated STR profiles show less than 15% of
> markers
> have diverged.
> --OR a BP for which NO specified population have been demonstrated to exist
> where the frequency is greater than 0.05% (P<0.05) and whose total male
> population exceeds 500 thousand individuals as defined geographically or
> ethnically.
> ===
> I guess it has slipped past me until now but this is apparently quite
> different to the definitions used by genetic genealogists, and quite
> different from what any native speaker of English would expect. I believe
> the normal common sense definition of "private SNP" is that the SNP
> actually
> been cross-checked and is NOT provisionally known about, but is actually
> known to be a relatively recent SNP that distinguishes only a very small
> group of closely related lineages.
> My ISOGG interlocutor has not provided much defence of this except one
> which
> was, frankly, ignorant. He said that he believes that with a few exceptions
> such as R-L21, there are hardly any SNPs which have been discovered which
> are likely to be private in the sense of being isolated to branch ends 500
> years or less. Obviously with WTY and 1000 genomes this is the opposite of
> the truth, and it means ISOGG are making an assumption which is least
> likely
> to be helpful in keeping tabs on the explosions of information which is
> coming our way. This inflexible way of working and thinking, demanding that
> reality should follow the commands of secret commitees, is not going to be
> well suited to the tasks ahead!
> That's why I broaden the forum of discussion, to see if there is some other
> counter argument out there! No other way of getting discussions seemed
> possible. I am posting on the GENEALOGY-DNA Rootsweb list, and on the ISOGG
> Yahoo group.
> One thing I am sure we expect from our organization's phylogeny pages, is
> some effort being taken to help us see which SNPs justify further testing
> and which do not. The definitions being used seem to be un-necessarily
> confusing? Even extremely misleading.
> PROPOSAL. Perhaps ISOGG need to define two levels of "provisional" or
> "under
> investigation" instead of wrongly using the word "private" to describe the
> least-studied provisional SNPs?
> I believe the word private implies that an SNP HAS BEEN STUDIED, and is no
> longer provisional.
> I wonder what others think?
> Best Regards
> Andrew
> -------------------------------
> 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

Jørgen K. Kanters, MD
Associate Professor
Laboratory of Experimental Cardiology
Dept. of Biomedical Sciences
University of Copenhagen
Blegdamsvej 3C
DK 2200 Copenhagen N
Phone: +45 35 32 74 02
Fax: +45 35 32 74 18

This thread: