DNA: Unexpected test results

In my last post we looked at autosomal DNA – what it is, and a little about how we can use the results of a DNA test to help us progress our family trees.  Before going any further let’s consider an important issue that should be borne in mind from the outset when deciding whether to test.

The fact is that DNA testing can reveal unexpected, uncomfortable information.  It is statistically unlikely, but we should be aware of the possibility when taking a test, and be prepared to deal with the consequences.  Our family trees may be meticulously researched, fastidiously sourced and referenced and the very model of genealogical good practice.  But DNA test results may reveal what birth registration and baptismal documents cannot.

These unexpected results might fall into two categories:

Non-paternity event or misattributed parentage event in your tree
The term ‘non-paternity event’ (NPE) is used in genetic genealogy to refer to any situation where DNA testing shows the biological father is not as expected.  It may be that the man we called ‘Dad’ is not in fact our biological father.  Or perhaps DNA testing of a close family member such as a sibling indicates that they have a different father.  It could even be that neither parent is biologically linked, or perhaps the DNA connections change a generation or so further back.  To allow for situations where the mother is not as expected, the term term ‘misattributed parentage event’ is often used as an alternative to ‘NPE’.

We need to keep this possibility in perspective.  FTDNA, one of the main DNA testing companies, assess the NPE rate at about 1-2% per generation. The International Society of Genetic Genealogy Wiki page on NPEs cites a number of studies, some of which have concluded that historical NPE rates were much higher than more recent times.  In other words, you’re more likely to find an NPE some generations back.  However with the passing of generations,  the cumulative impact of these per-generation statistics is significant.

When faced with evidence of an NPE it’s important not to jump to conclusions.  The above-mentioned ISOGG page sets out a whole range of possible explanations for how this might have come about, including:

  • Illegitimacy outside marriage
  • Infidelity within marriage
  • Remarriage, in which the child takes the surname of the step-father, perhaps after the untimely death of the biological father (e.g. in wartime)
  • Rape
  • Adoption, including hidden or informal adoption
  • Intentional or unintentional baby swap
  • Unintentional embryo swap
  • Sperm donation

It is entirely possible that your DNA test has uncovered a well-kept secret – even, in the case of unintentional baby/ embryo swap, a fact unknown even to the parents.

Another tester’s NPE or parents unknown that impacts on your family or ancestry
More likely is the possibility that your own DNA results will be absolutely in order, but that another tester’s are not.  Although my own DNA test has uncovered no direct NPEs, in my wider tree I’m aware of nineteenth century illegitimacy before marriage; infidelity or possibly rape (a child conceived while the husband was away at war); and informal adoption and remarriage following the death in WW1 of the father, resulting in the registering of the child in the name of the second husband.  An unexpected high match between you and another tester may indicate that their biological father is someone connected with you.  More likely yet, in my experience, is that some of your matches will have tested specifically in the hope of finding their own missing father, grandfather or even (if they are adopted) both biological parents.  You may find that quite a few of the people who match you but have no tree linked to their DNA results will fall into this category.  I will always do whatever I can to help any of my good matches in this situation.

How does the DNA reveal this?
Think back to the Shared centiMorgan Project chart in my last post.
We noted that we can expect to share 3485cM with a parent, but that the parameters are 2376-3720.  If my DNA results reveal a match of that size then there is no doubt: either I am that person’s parent or they are my parent.  Conversely, if my father or a close family member on his side (my aunt, uncle, cousin) has tested but we share no DNA at all, then either this is not my biological father or he is not a full sibling to this family member.  If, on the other hand, a completely unknown person matches me at half sibling or at cousin level, then one of my parents or one of my aunts or uncles had a child about whom I had no knowledge.

It is unlikely to be as straightforward as that. More often, our match will be more distant – the amount of shared cM will be smaller.  What we have to do is work with what we already know, gradually homing in on where we connect.  For example if I don’t match with A’s mother, then even though A has no paternal tree at all we can assume I do match with the father.  If A matches with my maternal cousin as well as with me, then we know it is my maternal side where we match.

In a sense, in this type of situation we turn the search on its head.  As outlined in my last post, the usual way to use DNA test results is to look at the amount of DNA shared (expressed as centiMorgans, the higher the number the closer the match) and then to look in our trees for our Most Recent Common Ancestor – Bingo! End of search!  Clearly, if our match has no tree and no knowledge at all of their birth family we can’t do this.  We can only progress if there are other people who match both of us.  Let’s say A is seeking their birth parents, and A and I match at around second cousin.  B and C also match the two of us but at around third or fourth cousin level.  If I can find in my own tree where B, C and I share an ancestor, there’s a good chance that A and I will have common ancestors one or two generations closer to me on that same line.  However, unlike me, A’s goal is not to identify which set of great grandparents they got that segment of DNA from.  What they want to know is the name of an unknown parent.  Therefore, if we can identify our own Most Recent Common Ancestor, they will now need to work forward from that, to identify their biological parent.  In the above example, on the basis of information available from our shared DNA, I can’t tell them that, but I can say that their unknown parent is probably the grandchild of these two specific people who are my own great grandparents, and that has narrowed down their search quite considerably.  It may take a while but eventually, as more people test we may have more testers matching both of us at a closer level, and we can continue to narrow down the possibilities.

Undoubtedly the above scenarios raise a number of ethical issues. We will consider them in my next post.


Edited August 2020
My posts about DNA are aimed at complete beginners and aim to provide information in manageable chunks, each post building on previous ones. Click [here] to read all of them in order, or to dip in and out as you wish. You’ll also find lots of resources and useful links