DNA  and Genealogy

How does DNA testing help in building a family tree

Our family had some excitement in 2010 when DNA testing convinced us that Keros’s brother Kish emigrated to Newfoundland, now a province of Canada.  These results are described in ‘KishDNA’.  At the same time we found that the DNA results showed that Les Caines and I matched Chris Keene with a genetic variation of 1 – see article ‘Chris Keene’.  Then at the end of the year another match surfaced with a genetic variation of 1 for a ‘Walter Pierson Dean’.  So what can we make of that?

I have a very limited knowledge of genetics and DNA laboratory testing, but from a genealogy point of view I have a good reference book entitled “DNA and Family History” by British author Chris Pomery.  Just a few extracts from this book has, for me, placed the Dean entry in its right context as described below.

Relevant quotes from the Pomery book followed by a discussion of how I see the quote applying to our family history follow:

QUOTE:    “…..Current academic studies cite an average rate of mutation for the average marker of 1 in 500 cross-generation transmissions of DNA, though some test firms believe the rate is closer to 1 in 350……”

DISCUSSION:           I understand this to mean that a mutation in our family line may have taken place as recently as between a current generation male and his father or as long ago as up to 350 to 500 generations ago.  If we assume 25 years per generation then the mutation may have occurred from any time from the present to some 87 to 125 centuries ago.  125 centuries ago takes us back to pre-historic times.  Even if the marker is a relatively fast-moving marker it could have happened a markedly long time ago.  And there is little chance of determining how long ago.

QUOTE:   “…..two people with unrelated surnames will virtually never find a documentable family tree connection even if their Y-chromosome haplotype is identical…..”

DISCUSSION:          Bearing in mind the 1 in 500/350 rates for mutation, if a mutation had not occurred for many generations, we might find a perfect match at 37 locations from 2 individuals with a common ancestor many centuries ago.  For genealogy and family trees we can only confirm relationships in relatively recent times when we can consult written records.  In relatively recent times, surnames of legitimate births are not likely to markedly change.  So a DNA match where the surname is different, I attribute either to a relationship many centuries ago, or to the 5% probability per generation of adoption or a birth where the father is not the husband of the mother.

QUOTE:   “…..a DNA high resolution test of 37 markers for two males with the same or similar surnames where the result is an exact match, then a link between the two is highly likely.  For a 1-step difference at 1 marker in the match, a link is quite likely……”

DISCUSSION:              These assessments are based on a large number of actual results in a company’s data base and while in an individual case linkage can be hugely variable, the best that can be done is to use a probability formula developed by geneticists.  Pomeroy has broadened probabilities meaningful to an amateur like me with his ‘highly likely’ and ‘quite likely’ assessments.

CONCLUSION:        I am in no doubt that DNA testing can help in identifying relationship within a family.  However, such relationship must be relatively recent –recently enough to match DNA records with written records which are meaningful in developing a family tree.  Go back too far and not only do written records be non-existent, but also names can hardly be recognisable.  From the above I am convinced that a link to Walter Pierson Dean is unlikely and any further research unwarranted.  I would certainly be interested in any contrary opinions.