The POMEROY GENETICS PROJECT #3
Description of the DNA Testing Programme
The lab performing the DNA tests was the Institute of Molecular Medicine at the
University of Oxford. The actual research was done by an MBioch. student, Chris Hardy, supervised by
Professor Bryan Sykes, Professor of Human Genetics.
DNA samples were collected using as small, stiff toothbrush supplied in kit form by the lab that is stroked
on the inside of the cheek and then posted back to Oxford for analysis.
After the analysis the DNA samples were destroyed. The only information retained are the records of the
markers in the DNA that were tested. The 14 markers used were: M9, 92R7, SRY 1532, YAP, DYS 19, DYS 388, DYS 389i
DYS 389ii, DYS 390, DYS 391, DYS 392, DYS 393, DYS 425 and DYS 426.
The Tests
The DNA test of the Y-chromosome consisted of two parts, an allele test and a microsatellite test.
The allele test reveals the haplogroup of each sample. This section of the Y-chromosome changes only very
slowly and with a known mutation rate, so it is a very useful way to distinguish between large numbers of samples.
Any member of the human population will belong to one of just a handful of haplogroups.
Indeed, most Europeans will belong to one of two extremely common haplogroups. The four markers tested were M9, 92R7, SRY 1532 and YAP.
The 51 POMs tested all fell into these groups which I have labelled Haplogroup 1 and Haplogroup 2.
The key fature about the haplogroups is that one can conclusively say that anyone in Haplogroup 1 cannot possibly
belong to a Haplogroup 2 family, and vice versa.
These distribution in the two haplogroups were then further subdivided by investigating another area of the
DNA known as microsatellites, in effect pieces of highly mutable 'junk DNA' that fill up most of the chromosome.
For the Pomeroy DNA programme the lab tested 10 microsatellite regions. The test result for each area of the
microsatellite is expressed as a number. The sequence of numbers set up by all the microsatellite areas put
together side by side is known as a haplotype. This can be thought of as a DNA signature.
If two tests record identical numbers at each and every part of the
microsatellite one can assume -- at least statistically -- that if they share the same surname that they share the same male ancestor.
The analysis gets more uncertain if the match is near but not identical. If the result in one of the 10 areas of the
microsatellite being tested is different by one unit from another testee this difference is described as a 'one-step mutation'.
The problem is that this one-step difference could indeed be the result of a random mutation of that
microsatellite in the gene. But it could also indicate that the gene has been formed by a completely
different male DNA, i.e. by another individual. One way to estimate the likelihood that the difference
in that one microsatellite is really a mutation from an existing haplotype is to compare the frequency with which it is
found in a control population. If the haplotype is very rare in the control population but seen
more than once in a small set of tests within a one-name group, then it is relatively likely that it is a mutation within
the group rather than resulting from an interloper's genes. Vice versa, if the
haplotype is seen only once in the set of testees but is a very common haplotype in the general human
population, it is less likely to be a mutation. Every option has to be thought of in terms of its probability.
The results of the haplotype analysis are charted in a phylogenetic network, a graphical way of
representing each set of microsatellite test results relative to their nearest neighbours.
Each node on the network -- which actually looks more like a grid or matrix than a traditional family tree -- is a
distinct haplotype. A node that is directly linked is a haplotype with a one-step difference in it.
The next node over would have differences from the original haplotype at two points of a single microsatellite.
When viewing the phylogenetic network charts later on, remember that the angle between the nodes is irrelevant, as is
the length of the line between them. What is important of the number of testees found per node (i.e. per haplotype)
and the number of nodes linking them together.
Summary of Key Results
51 DNA samples were successfully tested of the 66 POMs that were invited to return samples. The
remaining 15 samples were either not returned, were damaged in transit or did not analyse correctly
when tested in the lab.
The surname breakdown was 4 Pomery, 9 Pomroy and 38 Pomeroy.
The 51 POMs sampled belong to families that have 291 known living adult males POMs in them.
These 291 people represent 36.7% of the total number of 794 adult males known at the time of the tests to be living in the UK.
100% of the 51 tested POMs belong to the two main haplogroups found in Europe,
a characteristic that they share in common with around 60-65% of the total male population of Europe.
Almost two-thirds of them (31) belong to the most common European haplogroup (Haplogroup 1) and
slightly over one-third (20) to the second-most common haplogroup (Haplogroup 2).
There are seven clusters of related haplotypes within these two haplogroups.
This suggests that there could have been at least seven separate originators of modern-day Pomeroys.
There was no modal or dominant haplotype that would strongly indicate a single common ancestor.
It is possible that, since the name dates back earlier than the convention of English surnames, it could have been
subject to more infiltation of other DNA material and to genetic mutation, both of which factors could hide a single origin.
Single origin sources of the Pomeroy & Pomroy surnames could not be identified. However, in the much smaller
Pomery group only four men were tested. One fell within a much larger group containing Pomeroys and with
its origin in Cornwall some three centuries ago; another Pomery tested fell within a single mutation of this
group. Both the main group and the single-step mutation are relatively rare haplotypes in the control population.
This cluster may therefore be the single origin of the Pomery name (with the other two Pomerys tested being more
recent surname mutations from genetic stock now labelled as Pomeroy).
The multiple origins result largely destroys any idea that most of us bearing the name today are in any way related
to the Norman family of Ralf de la Pommeraye. Perhaps six of the 51 testees are, however, potentially related to
the ennobled Irish Harberton family which traces its line back to the Pomeroys of Ingesdon in Ilsington, Devon,
and thence to the Norman family.
None of the other five testees have identical DNA records with the aristocratic line though all bar one trace their
links back to Devon between 1715 and 1912.
The results have provided us with valuable clues where to look for documentary evidence for links that we have not yet
confirmed via the usual documentary sources. A total of nine DNA haploypes were found more than once in the 51 samples, and
these nine haplotypes accounted for 30 our of the 51 testees.
It is highly likely that individuals within each of these groups share a common male ancestor in the historical past.
The largest group had 7 members (of the 51 tested), two groups had 4 members, and there were three groups each with
three and two members.
Next Phylogenetic Networks for 31 POMs in Haplogroup 1
Last updated: 9 July 2001 |
