Personal genomics question: am I related to my 5th cousins?

There are some personal genomics questions I get over and over via email. I thought I would post an answer so that Google could pick it up.

One of them usually has do with if someone is “really” related to someone who comes up as a 5th cousin on a DTC service. What does “really” mean?

Graham Coop has done the formal work to show that it’s highly likely all our genealogies intersect some point in the recent past. Several years ago in his paper with Peter Ralph, The Geography of Recent Genetic Ancestry across Europe, they used genetic data to infer just how closely European lineages coalesced with each other over the past few thousand years.

So yes, you are related (though that doesn’t mean you have matching genetic segments).

But that’s not the question people are really asking about. They are asking, does this DNA match increase the probability that I’m somehow related to this person?

In general, I think not. For example, I regularly get queries from South Asians about distant matches with Europeans. Does this mean they are European? No. I think what it means is:

1) There are lots of Europeans in the database, so a false positive match is likely to be European, even if you are non-European.

2) At short genetic distances, the segments are really mostly some sort of false positive.

One thought on “Personal genomics question: am I related to my 5th cousins?

  1. If you have phased genomes, you can cut down the number of false positives. I was able to do this on GEDMATCH and find an exotic (for me) cousin in Indonesia.

Comments are closed.