DNA Testing

[MelPack]

It is a matter of some regret but I have become disillusioned with many aspects of the MoD's SPVA unit's and the CWGC's handling of the relatives' side of the Fromelles Project.

One cause of that disillusionment is that no clear guidance has been issued to the relatives about the most suitable candidate that they should be seeking to locate both within their immediate families and amongst their more distant relatives for the DNA test.

The only concession that has been made hitherto in that direction is a brief exposition posted on the CWGC's Fromelles website, the key second diagram of which appears to have been so mangled in the process of uploading/editing that it has been rendered meaningless as an explanation for the transmission of YDNA and MtDNA from one generation to the next.

Currently the relatives are being asked to 'populate' their family trees with as many people as possible and to pursue female relatives in particular pending individual advice that they will eventually receive from a DNA expert.

I have received emails from a number of relatives who are completely befuddled as to which other relatives to search for and why. In the absence of clear guidance on this matter from the SPVA, one of our colleagues from the genealogy project, Michelle Leonard, has kindly settled the following explanation.

Mel

[MelPack]

DNA TESTING IN RELATION TO FROMELLES

There are two types of DNA that are relevant for testing purposes - Chromosomal or Nuclear DNA and Mitochondrial DNA. Assuming that viable samples of either type can be recovered from the remains buried at Fromelles via bone tissue or, more likely, vascular tooth pulp then there are three types of DNA tests that could be conducted:

1) Mitochondrial or MtDNA Test

Mitochondrial DNA http://damian.peterson.net.nz/tag/mtdna/ (MtDNA for short) is a non-chromosomal DNA that exists in the body in vastly greater quantities and is much more durable than chromosomal DNA. It is one of only two types of DNA that are passed down virtually unchanged over hundreds of years making it ideal for testing purposes and it has, therefore, been the preferred testing method when discoveries of ancient remains have been made in the past. It is for these reasons that relatives are being urged to locate potential candidates that can provide a sample of this type of DNA.

MtDNA is only passed from one generation to the next from mother to child; it is passed to both sons and daughters but only the daughters can pass it on in turn to their children and so on (e.g. a soldier and all of his brothers and sisters who shared a common mother would have inherited exactly the same MtDNA from that mother but the children of the males would not pass on this MtDNA because their children would inherit the MtDNA of their own mothers while the children of the females would pass it on to their children and these same strand MtDNA lines would perpetuate until the interjection of a male brings each individual one to an end.)

This means that only certain female relations of the soldier could have transmitted the same MtDNA he possessed to future generations and, therefore, for a living descendent to share his MtDNA they have to descend from his mother, maternal grandmother, maternal great grandmother and so on down an unbroken female line (matrilineal descent) whether they themselves, as the current end point, are male or female.

[MelPack]

The following living relatives would be suitable for this type of test:

Nephew/Niece (must be the son/daughter of a daughter of the mother of the soldier)*

Great Nephew/Niece (must be the son/daughter of a daughter of a daughter of the mother of the soldier)*

Grt Grt Nephew/Niece (must be the son/daughter of a daughter of a daughter of a daughter of the mother of the soldier)*

First Cousin (1xRemoved) (must be the son/daughter of a daughter of a daughter of the maternal grandmother of the soldier)**

First Cousin (2xRemoved) (must be the son/daughter of a daughter of a daughter of a daughter of the maternal grandmother of the soldier)**

*You may think it would have been easier for me to say "sister of the soldier" in this instance but either one of the soldier's parents may have re-married and the soldier could have half sisters who are the daughters of his father and therefore would not share the same MtDNA. So then why didn't I just say "full sister of the soldier" in that case? The answer is that equally he could have half sisters with whom he only shares a mother and in that case these half sisters would share the same MtDNA and are just as relevant as full sisters would be for the purposes of MtDNA testing so the main rule to follow is that the sister must be the daughter of the mother whether she is a full sister or a half sister.

**The same caveat exists here; the maternal aunt must be the daughter of the soldier's maternal grandmother and, if so, she will still share the same MtDNA as the soldier whether she is a full aunt or a half aunt but if she is the daughter of the soldier's maternal grandfather she will not share it at all.

[MelPack]

It cannot be emphasized enough that it is not sufficient (as implied by the SPVA) to find a living female descendant of a soldier’s sister, maternal aunt or maternal grandmother and so on; the living descendant must come from an UNBROKEN female line that stretches back to at least the soldier's mother or maternal grandmother. In practice, therefore, this means that not every female on the maternal side is suitable for MtDNA testing (e.g. the daughter of a SON of a sister of the soldier cannot carry the same MtDNA as the soldier since her father could not pass it down to her and she will have inherited the MtDNA of her own mother who is not blood-related to the soldier in any way.) It is important to remember that as soon as a male enters the unbroken line the MtDNA of that line is effectively dead as it will switch to the MtDNA of the female partner of that male.

It must similarly be stressed that the living descendant does not necessarily have to be female; a living male descendent can be at the end of the unbroken female line just the same as a female one can since MtDNA, unlike the Y-Chromosome, is passed down by a mother to BOTH her sons and her daughters (e.g. a great nephew that is the son of a daughter of a sister of the soldier carries the same MtDNA as the soldier since the line above him is an unbroken female one that extends back to the soldier's mother and as he will have inherited his mother's MtDNA exactly the same as his sisters would he is just as suitable as them for MtDNA testing.)

[MelPack]

Neither of these sequences would work:

1) Full sister of a soldier to her daughter to her GRANDSON to her great granddaughter (the intercession of the grandson means the great granddaughter possesses the MtDNA of the grandson's wife)

2) Full maternal Aunt of a soldier to her daughter to her granddaughter to her GREAT GRANDSON to her Grt Grt granddaughter (again the intercession of a male brings an end to MtDNA inheritance from this line.)

Conversely these sequences do work:

1) Maternal half-sister of a soldier to her daughter to her granddaughter to her GREAT GRANDSON (as the great grandson is this time at the end of the unbroken line he will share the same MtDNA as the soldier.)

2) Full sister of the maternal great grandmother of the soldier to her daughter to her granddaughter to her great granddaughter to her Grt Grt granddaughter to her Grt Grt Grt granddaughter (although this Grt Grt Grt granddaughter is only a third cousin twice removed of the soldier she shares the same MtDNA as him albeit with a few more slight mutations that have crept in over the generations than that his sister would have shared with him but, even although their genes will not be close in any other way, this should suffice for MtDNA testing.)

THIS CHART shows the line of inheritance needed for MtDNA testing.

http://www.kerchner.com/mtdnachart.htm

[MelPack]

2) Y-Chromosome DNA Test

The Y-Chromosome is found solely in males and is, therefore, the sex-determining chromosome for men. It is paternally inherited and, as such, is transmitted from one generation to the next exclusively from father to son. Additionally, as it is the other DNA material that is passed down virtually unchanged over hundreds of years and, as such, is very suitable for DNA testing purposes. This test, however, can only determine a common ancestry between two males and, therefore, all female relations are excluded from it.

Since the soldier inherited his Y-Chromosome from his father and his father in turn inherited the same Y-Chromosome from his father any male relation of the soldier on his paternal side who is descended directly from the soldier's father, paternal grandfather, paternal great-grandfather and so on down a continuous unbroken male line will, therefore, carry the same Y-Chromosome as the soldier himself.

[MelPack]

The following living relatives of a soldier would be suitable candidates for this type of test:

Son

Grandson (must be the son of a son of the soldier)

Great Grandson (must be the son of a son of a son of the soldier)

Nephew (must be the son of a son of the father of the soldier)*

Great Nephew (must be the son of a son of a son of the father of the soldier)*

Grt Grt Nephew (must be the son of a son of a son of a son of the father of the soldier)*

First Cousin (1xRemoved) (must be the son of a son of a son of the paternal grandfather of the soldier)**

First Cousin (2xRemoved) (must be the son of a son of a son of a son of the paternal grandfather of the soldier)**

*As with the MtDNA testing I have not used "brother of the soldier" as the soldier may have had half brothers with whom he only shares a mother and in those cases that brother and certain of his descendants would be suitable for MtDNA testing but not Y-Chromosome testing so for the purposes of this test the brother must be the son of the soldier's father.

**Similarly the soldier may have paternal half uncles who are solely the sons of his paternal grandmother and these half uncles will neither share MtDNA nor a Y-Chromosome with the soldier. It is important, therefore, to remember that to be suitable for Y-Chromosome testing a paternal uncle must be the son of the soldier's paternal grandfather and this is especially relevant if he is a half uncle..

[MelPack]

As with the MtDNA test even more distant relations to the soldier could provide a viable sample for this type of test; e.g. a male descendant of a brother of the paternal grandfather or even paternal great grandfather would share the same Y-Chromosome as the soldier provided that they are a part of an unbroken male line (patrilineal descent). THIS CHART shows the line of inheritance needed for Y-Chromosome testing.

http://www.kerchner.com/ydnachart.htm

The difficulty, however, with the Y-Chromosome is that it is more perishable than MtDNA and exists in much smaller quantities (i.e. there is only one Y-Chromosome per cell in the human body yet there are between 100 and 10,000 MtDNA per cell). When you factor in the fact that DNA deteriorates over time then it may be difficult to extract Y-Chromosome DNA from the remains at Fromelles and this is the reason why relatives from the female line are being sought even though they may be far more distant relations than others who have come forward.

[MelPack]

3) Autosomal DNA Test

This is a chromosomal DNA test that uses DNA profiling and is applied to establish the commonality of genes between close relatives within one or two generations of each other. The most common way in which this test is applied is in cases of disputed paternity and since women do not have a Y-Chromosome it is especially useful in determining the paternity of a female child. Through an autosomal test it is possible to establish how many shared chromosomes two people have and even pinpoint the exact relationship between them but this becomes more difficult the more distantly related the two people involved are as they will share a smaller percentage of genes.

Each person inherits half of their 46 chromosomes from their mother and the other half from their father so they should share 50% of the same genes as a parent or sibling and approximately 25% of the same genes as a grandparent. After this point the percentages lessen accordingly: Full first cousins share approximately an eighth or 12.5% of the same DNA, first cousins once removed share 6.25%, first cousins twice removed share 3.125%, second cousins share 3.125% also, second cousins once removed share 1.563% and second cousins twice removed share 0.781%. This is a good illustration of why autosomal testing can only be conducted with close relatives and THIS CHART is very helpful when trying to work out the percentages that pertain to different relations.

http://tevern.us/genealogy/Cousins/index.htm

[MelPack]

This particular test, therefore, could be used to establish a family link between a soldier and a close generational living relative who neither shares a Y-Chromosome or MtDNA with him. The relationships that this would encompass in regards to the Fromelles soldiers are:

Grandson (this would usually be applied to the son of a daughter of the soldier as the son of a son would be suitable for a Y-Chromosome test but it can be conducted on both)

Granddaughter (this is the only test applicable to a granddaughter of the soldier no matter whether she is the daughter of his son or the daughter of his daughter)

Niece (this would usually be applied to the daughter of a brother of the soldier as the daughter of a sister would be suitable for an MtDNA test)

Note this list does not include "Nephew" even although a nephew would be just as suitable as a niece for autosomal testing; I have not included this category as a nephew via a brother will carry the same Y-Chromosome and a nephew via a sister will carry the same MtDNA as the soldier.

The difficulty with this type of test is the same as that for the Y-Chromosome DNA test in that chromosomal DNA as a whole is more perishable, exists in much smaller quantities and is harder to extract from old bones than MtDNA.

[MelPack]

In conclusion, it is important to think outside the box with DNA and not to discount distant relations as, although a living niece or nephew is a very close generational relative and may well be the perfect candidate for DNA testing, some instances may arise in which they are not the best candidates (e.g. If it is not possible to extract chromosomal DNA from the remains of a particular soldier then the only possibility for identification will be through MtDNA. In that case a living niece who is the daughter of the soldier's brother will not be the best DNA match since she does not share MtDNA with the soldier. A living second cousin twice removed, however, who is clearly much less closely related but is the son of the daughter of the daughter of the daughter of the full sister of the soldier's grandmother and, thus, does share the same MtDNA as the soldier would be a much more viable match in this instance.) Also be careful about which maternal female relatives to trace - as I have pointed out they are not all of equal value in DNA terms.

Finally I must stress that until we are informed of how the extraction procedure is proceeding and which DNA testing methods can be used on the remains of each individual soldier this is all informed guesswork that is designed to cover all eventualities and lead us to the best chance of fulfilling the overall objective; getting each of these soldiers a named grave.

[Fedelmar]

http://www.smh.com.au/world/all-remains-ex...90809-ee8e.html

No media release from the Australian Minister as yet.

6.45am W. Australian time.

Bright Blessings

Sandra

[Fedelmar]

Minister for Defence Media Mail List

------------------------------------------------------------------------

THE HON. GREG COMBET MP

Minister for Defence Personnel, Materiel and Science

Monday, 10 August 2009

035/2009

DNA TESTING OF FALLEN AT FROMELLES TO PROCEED

Greg Combet, Minister for Defence Personnel, Materiel and Science, today announced that full analysis and matching of DNA from Australian and British World War One soldiers discovered in France will proceed.

Today's announcement follows the successful conclusion of a pilot study to evaluate if the DNA present in the remains found at the Fromelles site could be used to identify our fallen soldiers.

"I am pleased to announce that the results of this pilot study have clearly demonstrated that viable DNA samples are able to be collected from a significant number of the fallen at Fromelles," said Mr Combet.

"The pilot study tested a cross section of the Fromelles remains with samples being taken from the teeth and bones of these remains. Samples were also taken from different parts of the burial sites.

"On the basis of the information provided by the study, the Australian and United Kingdom Governments have agreed that a full DNA testing programme will be conducted.

"The delicate condition of the remains, the soil surrounding them and high water table has made the extraction of DNA difficult. Therefore it is also important to note that the chance of identification of each solider still remains low," said Mr Combet.

"Both the Australian and United Kingdom Governments remain committed to identifying as many of the fallen as is possible.

"DNA provides another piece of the complex identification puzzle. But as important as this is, it is just one of the many techniques that will be used to try and identify these soldiers.

"To increase the chance of success, identification of individuals will also rely on a combined picture of anthropological, archaeological, historical and scientific data."

Over 1300 descendants of Australian soldiers who died in the Battle of Fromelles with no known grave, have registered with the Australian Army to offer their DNA to match with the remains.

"Descendants who are identified as being the best chance of a DNA match to a soldier who died at Fromelles will be contacted directly in the coming months."

Given the public interest in this historic project the Government is currently considering the release of further information relating to the DNA testing and pilot report.

If people believe that they are relatives of a soldier who died at the Battle of Fromelles, they are encouraged to contact the Australian Army on (free call) 1800 019 090 or visit www.army.gov.au/fromelles to register their interest.

7.55am W. Australian time.

[Auimfo]

Beat me by a whisker Sandra!!

Cheers,

Tim L.

[MelPack]

Tim & Sandra

Whatever the rarity of the debate of the rib versus the femur, I think that it was always going to be the case that vascular tooth pulp would be the primary source of viable DNA samples for the bulk of the remains.

I hope that Greg Combet does release the pilot report so that the number of remains tested and the success rate for extracting viable samples is in the public arena.

I also hope he stands by the indication that further information on the DNA testing programme will be released particularly in guidance for the relatives. It would set a useful example for the British MoD to follow.

Mel

[dfaulder]

>><<Each person inherits half of their 46 chromosomes from their mother and the other half from their father so they should share 50% of the same genes as a parent or sibling and approximately 25% of the same genes as a grandparent. After this point the percentages lessen accordingly: Full first cousins share approximately an eighth or 12.5% of the same DNA, first cousins once removed share 6.25%, first cousins twice removed share 3.125%, second cousins share 3.125% also, second cousins once removed share 1.563% and second cousins twice removed share 0.781%.>><<

Mel,

I would like to know the reference for this as I (from a non specialist scientific background) do not see it as quite this simple.

Having inherited half my chromosomes from my father and half from my mother, if I was to have children, I would pass half my chromosomes on to my child, but, it is extremely unlikely to be half of those I inherited from my father and half of those that I inherited from my mother (indeed it is mathematically impossible as 23 does not divide by 2). That would make genealogy far too easy.

Imagine my father had a pack of 46 cards (with a blue back) and my mother another pack of 46 cards (with a pink back). Each deals me 23 cards, to give me 46 - which I then shuffle - before I deal 23 to my child. My child might get 0 that originated from my father and 23 that originated from my mother (or vice versa or any combination in between).

My understanding is that this sort of DNA testing takes us into the realms of probability with the classic bell shaped curve giving the likelihood of passing on a particular proportion of a given ancestor's chromosomes. Whilst the extreme cases (e.g. my child getting none of the chromosomes that I inherited from my father) are unlikely - think of dealing the mixed pack of cards, it is possible that full siblings have no chromosomes in common. I think the "approximate" in the above quote is what statisticians call an un-typical mean (i.e. a point right on the tip of the bell shaped curve - in theory the mean but not especially likely to exactly occur). This is why the DNA testing services offered by some genealogical companies have to express a confidence limit (i.e "from autosomal examination of the two samples we are 90% confident that they come from cousins, x% confident that they are from second cousins, etc.").

The other types of testing do not (as far as I know) suffer from this uncertainty.

David

[MelPack]

David

You are absolutely right about probability and randomness. There is no mathematical absolute and that is why the term approximate was used as a shorthand for all the complexities involved in the transmission of chromosomes.

Take a father and daughter. The daughter will have half of the father's 46 chromosones so it should be fairly easy to get a match. Which 23 of her father's chromosones she received, however, is completely random so she could have received any of the 46 he inherited from his parents. So any sibling, unless they are an identical twin, will have received a completely different mix of 23 chromosones from the same 46; hence why siblings can be so similar and also so different.

In the extremis that you cited, I suppose it's possible one sibling could inherit all 23 of the 46 chromosones their other sibling didn't inherit but I think that would be hugely rare and identical twins are the only people who get the same 23 from both parents thus having identical DNA. These chromosomes also mutate slightly from person to person with the exception of the Y-chromosone and MtDNA which must make it a little more difficult for the scientists. So if any two brothers can perhaps have around 50% of the same chromosomes from their father then a paternal uncle may share around 12 chromosones with a niece (ie 25% share) and it would be around the same chances for a grandfather and grandaughter.

It is, as you say, about probability.

Mel

[Michelle79]

Hi David,

I wrote the above DNA explanation and I also am not from a scientific background but have done entensive reading on DNA transference; the reference comes from HERE and I agree that it is an overly simplistic way of explaining the possible percentages of genetic relationship since it doesn't properly illustrate the level of randomness involved in the transfer of autosomal chromosomes. You are entirely correct in your summation of chromosome transference and the fact that these percentages are a very vague approximation since the only fact that can be stated with absolute certainty is that you inherit exactly 50% of your DNA from your father (i.e. 23 of his 46 chromosomes of which 44 are autosomal and 2 are sex-determining) and the other 50%, with the exact same breakdown, from your mother. Which 22 of your father's 44 autosomal chromosomes you inherit and which 22 of your mother's 44 you inherit, however, is completely random. As far as the two sex-chromosomes are concerned you inherit one from each parent; which sex chromosome you inherit from your mother is also random since a female possesses two X-Chromosomes and can, therefore, only pass one of those down but which one you inherit from your father is not random at all since he possesses one Y-Chromosome and one X-Chromosome so if you are male you must have inherited his Y-Chromosome and if you are female you must have inherited his X-Chromosome (i.e. since you are male you definitely inherited your father's Y-Chromosome and one of your mother's two X-Chromosomes but if you were female you would have definitely inherited your father's one X-Chromosome and one of your mother's two X-Chromosomes).

The percentage of DNA each person shares with any relation other than a parent is unique and highly variable due to the way DNA is passed down in this mixture of chromosomes (i.e. it is highly possible you may have inherited a higher percentage of chromosomes from your paternal grandfather than you did from your paternal grandmother as, although it can be categorically stated that you received 50% of your DNA from your father, that does not necessarily mean you will have inherited exactly 25% from each of his parents.) In the case of parents the existence of obligatory genes mean a positive result can be achieved without the need for statistical probabilty but for all other relationship tests a Likelihood Ratio is used and the more distant the relationship the more approximate the percentages are likely to be. You are also correct to note that Y-Chromosome and MtDNA testing do not have such probabilities attached to them due to the fact they are passed down virtually unchanged for hundreds of years.

Since only the 50% figure for parental inheritance is entirely accurate this means even the approximate 50% you share with a sibling is highly variable and the actual relatedness could theoretically range from 0% to 100% so it is statistically possible for a person to share more DNA with a sibling than they do with a parent. It is, as you say, conceivable, although virtually impossible, that one sister could end up with an entirely different set of 44 autosomal chromosomes to another sister but two full sisters have to share at least one chromosome as they both receive the exact same X-Chromosome from their father since he only has one X-Chromosome and the same applies to full brothers sharing their father's Y-Chromosome. A sister and a brother, however, could hypothetically share no chromosomes at all if they inherited the exact opposite sets of autosomal chromosomes from their parents and the boy then received the father's Y-Chromosome and one of the mother's two X-Chromosomes as his sex determining chromosomes and the girl received the father's X-Chromosome and the mother's other X-Chromosome but they would still share MtDNA. You would be hard-pressed to find this pair though and indeed I have seen no evidence that any instances of this phenomenon have ever been documented. Conversely siblings could share more than 50% of chromosomes with each other but the only set of siblings that you can state a percentage for categorically are identical twins who share 100% of all their chromosomes; if non-identical twin siblings were found to share 100% of their DNA then I would suspect some radical scientist had succeeded in an attempt at human cloning as the odds on this occurring are less than one in a trillion (see HERE) and, once more, no examples that I know of have ever been uncovered. In reality I would expect most siblings to share somewhere between 40%-60% of chromosomes with 50% as a decent guide average and those who share significantly more or less to be relatively rare.

I included the approximate percentages in my explanation above as I felt they were useful as a rough guide for relatives to see how unlikely an autosomal match beyond the grandparent generation is and I did not expand on it as I did not want to make the explanation even longer or deviate too much from the main objective of helping relatives understand which distant relations may make the best DNA matches. I have since altered this paragraph slightly on my master copy, however, to include more of an emphasis on the high level of approximation involved as after re-reading it I felt, like you, that this needed to be made clearer but I forgot to ask Mel to alter it on here so apologies for any confusion. I have read a few articles which state that although autosomal testing is currently not possible beyond the grandparent generation scientists are currently working on broadening its scope but I'm sure, as you say, that would simply create a set of circumstances in which we would find even lower percentages of probability than we currently have for closer generations.

Mutations also have to be taken into consideration but I won't go into those complexities here. Ultimately I think autosomal testing will not be a highly used method in relation to Fromelles and, considering the latest press releases, I feel it is more and more likely MtDNA will be the main DNA testing method employed.

Regards

Michelle

Edit: I see Mel has replied with an excellent explanation above while I was writing this anyway!

[MelPack]

Just a brief update - where there are appropriate donors in a family tree then a maximum of two relatives are being sought for YDNA and two for MtDNA testing per soldier by the MoD.

Mel