Distinguishing haplogroups from each otherDetermining your haplogroup is not a streightforward task. There may also be some uncertainty connected to your reported haplogroup from the DNA testing lab. I have taken the liberty to repost parts of a letter originally written by DNA guru Prof. Bryan Sykes, Ph.D. In his letter, Dr. Sykes explains the process of distinguishing haplogroup U4 (Ulrike) from haplogroup H (Helena). As we learn more about these markers, we're better able to sort people into groupings. The letter is quite technical:
From: "Annie, The WritingTeacher"
Subject: [DNA] Letter to me from Bryan Sykes on distinguishing H from U4 with 356
Date: Thu, 05 Sep 2002 09:29:59 -0700
David tells me that on your original certificate, issued in August 2000, you were placed in the clan of Helena but that when you were sent a replacement you had become a daughter of Ulrike instead. Of course your actual DNA sequence hadn't changed, but the assignment of you clan had. It may help if I explain how that is done.
Clans are defined by a mixture of two sorts of genetic markers, the variants in the control region sequence and the variants at a number of other sites around the mtDNA molecule - now generally called SNPs (short for Single Nucleotide Polymorphisms). These are usually designated as +4643Rsa1 or +11329Alu1 etc as you have pointed out in your messages.
However, that would put the price to customers up hugely because each one of the SNPs had to be done separately - although I know that Oxford Ancestors are looking into offering this service.
But even that would not guarantee completely accurate assignment in every single case. Only sequencing the entire mtDNA circle of sixteen and a half thousand bases at astronomical cost would do that - though even that would not be any good unless you had at your fingertips a database of thousands of other complete sequences with which to compare it and only a handful have been completely sequenced to date. Also, as more work is reported, the evolutionary networks will change.
What I am getting at is that no system is foolproof. The Oxford Ancestors service, to keep it affordable, only sequences the control region. Then the sequence is compared to a database which holds other sequences which have been examined for SNP variants. If the customers control sequence matches up with one of these then it is assigned to the same clan. In other cases, where there is not an exact match, the database is searched for close matches or sites which are characteristic of particular clans. In the case of your sequence which has variants from the reference sequence at 189, 356, 362 (we delete the 16 prefix for HVS1) two of the three variants are quite unstable - that means they can mutate back and forth. Position 189 is one of the least stable of all and 362 is not very far behind. Position 356 is far more stable and is also characteristic of clade U4, whose clan mother is Ulrike. However, it is not completely stable and does crop up in other clans - one of which is Helena.
So the sequence 189, 356, 362 could be in the clan of Ulrike mutating at the unstable positions 189 and 362 away from the U4 root sequence of 356. Or it could be in the clan of Helena with a rare variant at 356 taking it away from 189, 362. One way of telling the two apart is to look at the variant at 073. This is actually in HVSII and not HVSI and that was a source of confusion in some of the email exchanges I have read. Oxford Ancestors doesn't do the 073 test, as you know, so the sequence was assigned on the balance of probabilities to Ulrike. I have now had a chance to compare the sequence to some new research data of my own from Britain in which we did do the 073 test and found five exact matches which carry A at 073, indicative of clade H. So I think you are probably correct and are indeed a daughter of Helena rather than Ulrike. This means that the mutation at 356 would have occurred on a Helena background rather than the 189 and 362 variants occurring on a Ulrike background.
That might explain why you were originally issued with a Helena certificate in August 2000. At that time, the service was being sent out from my laboratory before Oxford Ancestors acquired its own premises. That means that whoever did that first assignment, and it may well have been me, did recognize the ambiguous nature of the 356 mutation in that particular sequence but that piece of information was not properly transferred to the new set-up - and that is my fault.
I must thank you for clarifying the assignment of this particular sequence. It is a changing field and your observation has helped it move on one more stage further. I am sure Dr Ashworth will want to issue a new Helena certificate. And of course, I hope you are pleased to have moved back to your original clan.
Bryan Sykes MA PhD DSc
Professor of Human Genetics University of Oxford
Link to original Usenet posting.