8/14/2019 The Russian Journal of Genetic Genealogy Comment on Geographic distribution and molecular evolution of ances

1/3

RJGGThe Russian Journal of Genetic Genealogy: Vol 1,1, 2009

35

ISSN: 1920-2997 http://rjgg.org All rights reserved

Comment on Geographicdistribution and molecularevolution of ancestral Ychromosome haplotypesin the Low Countries byGerhard Mertens and HugoGoossens

Guido J Deboeck

Received: Jule 4 2009; accepted: Jule 4 2009; published Jule 8 2009The associate editor for JOGG, Arlington, VA in collaboration with several

others who reviewed this paper for an earlier submission to JOGG.E-mail reactions to gdeboeck@mac.com

1

Dear Editor-in-Chief

In Geographic distribution and molecularevolution of ancestral Y chromosome haplo-types in the Low Countries by Gerhard Mertensand Hugo Goossens, (published in number 1of volume 1 of the RJGG), the authors presentthe results of an analysis of a Y chromosomedataset for individuals with genetic roots in theBenelux.

Many historians, familiar with the history ofthe Low Countries, as well as many countrymenof the authors, will be surprised (and possiblybe dismayed) about the geographic distribu-tion that was created and adopted in this paper.Since this geographic distribution was the basisfor all comparisons of Y chromosome haplo-types, we (a number of countrymen and historybuffs) have serious concerns with regard theanalysis of the data on which the conclusionsare based. The paper is too short with regardto the history of this part of the world, which is

necessary to explain some of the results.The authors wrote that "in order to be in-

cluded in the study, the place of birth of the old-est known ancestor had to be located in the ter-ritory of present day Belgium, the Netherlands,Luxemburg or the North Department of France".However, to make meaningful comparisons be-tween Y chromosome haplotypes, the authorsconstructed artificial "regions" based on their

own "geographical, historical and cultural logic".Unfortunately this logic is not provided in thepaper. We respectfully submit that these arti-ficial groupings can hardly be justified on thebasis of established geographical, historical andcultural history of the Low Countries.

This paper uses data from ancestors bornbetween 1200 and 1922 (with 75% of casesbefore 1739) from locations in present dayBelgium, Netherlands, Luxembourg, and theFrench Department du Nord, grouped in sev-eral "regions". Some groupings e.g. Brabant,Limburg, Holland, Flanders are obvious. Howev-er, the distribution of provinces over the NorthProvinces and the East Provinces is not obvi-ous. To the extent that geography is the maindeterminant of common characteristics thepresented groupings are appropriate. The dif-ficulty arises, however, because over time sev-eral of these locations belonged to principalities,or moved between principalities of which thegroupings have no relationship to the regions to

which they are assigned for this study.For example, part of Overijssel was in the

14th century part of the 'bishopric of Utrecht',after having been part of Gelderland. Gelderlandis grouped as part of the East Provinces whileUtrecht is grouped as part of the region Holland.Some ancestors of Overijssel may, however,have more common characteristics with thoseof Holland than with the East Provinces. Also,during its early history Overijssel included muchof modern-day Drenthe. Therefore, some olderancestors born in Overijssel may have more

common elements with ancestors from the NorthProvinces than with those of the East Provinc-es. So, the geographic delineations adopted inthis paper may not be appropriate for some an-

1

8/14/2019 The Russian Journal of Genetic Genealogy Comment on Geographic distribution and molecular evolution of ances

2/3

8/14/2019 The Russian Journal of Genetic Genealogy Comment on Geographic distribution and molecular evolution of ances

3/3

RJGGThe Russian Journal of Genetic Genealogy: Vol 1,1, 2009

37

ISSN: 1920-2997 http://rjgg.org All rights reserved

packages and the results are interpreted in thecontext of the history of this region. However,there are a number offlaws both in the use ofthe software and the interpretation of the re-

sults. Specifically,The term "ancestry informative markers" is

mainly used for loci showing polymorphism inone population but almost not in other popula-tions. This term is therefore not appropriate forexplaining the increased frequency of certain al-leles in populations. This high frequency couldhave resulted from genetic isolation and a lowpopulation size. Furthermore, the out-of-Africamigrations that predate the arrival of Homo sa-piens in Europe has also resulted in a decreaseof genetic diversity (genetic bottleneck) with a

result that the non-African populations show dif-ference in genetic diversity with respect to theAfrican populations.

Usually when Fst values are determined it iscommon practice to determine the significanceof the differences by doing permutation tests.It is therefore difficult to see if there are anysignificant differences (except with the Africanpopulation) among the groups from the Beneluxfrom the figures in table 2. Furthermore, nega-tive values indicate no difference at all betweenthe data. It is unclear from the paper if thishas been taken into account for calculating the

Neighbour Joining (NJ) tree.The NJ tree shows some results that con-

trasts with publications concerning the Dutchpopulation by De Knijff et al. The largest differ-ences were seen between the Northern provinc-es and the Southern while in this paper it is anEast-West (East-Holland) difference. It is diffi-cult from the data in the paper to identify is thisdifference is due to a problem in the analysisperformed with the genetic software programsor is due to the sample set.

The network produced in figure 5 is prob-

ably not correct, which is also substantiated bythe remark of the authors that "if median vectorsare included the tree becomes very cluttered".This is mainly due to insufficient knowledge ofthe analysis of Y chromosome haplotypes withthe network software by the authors. Not all Y-STRs can be used for this analysis and one mustincorporate also differences in mutation rate inorder to get a network that reflects the evolu-tionary history of the haplotypes. In addition,one should make use of the haplogroup desig-nation to explore in more detail the evolutionaryhistory and to come to meaningful conclusionswith regard to the history of the populationsanalysed. This can be done by constructing phy-logenetic networks for each haplogroup or to

include the haplogroep designation in the hap-lotype. The recent paper by King and Jobling(Molecular Biology and Evolution 26: 1093-1102, 2009) could be helpful for this analysis.

Therefore, any conclusions (historical and an-cestral haplotypes) obtained from the networkin figure 5 is premature and should be regardedas not scientifically substantiated.

In a nutshell, the historical context of whatthe Low Countries went thru is missing andhence makes it more difficult for readers notfamiliar with this part of the world to grasp thecomplexities the paper is trying to unravel. Thepaper also omits any comparisons with GreatBritain, France, Germany which are the direct

neighbors of the former Low Countries. Thereare a number offlaws in the use of the freelyavailable software, not to mention wrong inter-pretations of results.

Sincerely,

Guido J Deboeck et al.Arlington, VA