Finding birth family by genetic means

Well, it's been a month now. I confess to not being good at waiting; in both leisure and work I tend to gravitate towards stuff I can get done myself at my own pace. So, to distract myself from the wait—and pursue my search from a different angle—I finally did something I've been thinking of for a while: I sent away a DNA sample for sequencing.

Genetics is an interesting new search tool for male adoptees, for a reason I'll explain below. I've known about the underlying science for some time (I happen to have a science background), but only recently learned of its potential for aiding an adoptee's search. I'll explain the science here and get into the details of what I ordered in a subsequent post.

The first thing to understand is that, roughly speaking, there are two kinds of DNA: recombinant DNA and non-recombinant DNA.

Recombinant DNA

Most of our DNA is recombinant. What does this mean? Every person has two copies of every gene, one from each of our parents. When our bodies produce eggs or sperm, most of the genes are shuffled around via a process called recombination, and one of them is thrown away. So half the genes from this parent are lost, but that's made up with with half the genes from the other parent, so it all adds up. However, the result is that we share half of our genes with a parent or sibling; 1/4 with a grandparent, aunt or uncle; 1/8 with a great-grandparent, first cousin, great-uncle, or great-aunt, etc.

The gist of this is that recombinant DNA gets so mixed up with each generation that it's hard to trace relationships beyond 2 or 3 steps. To use genetics to identify a birth relative, you'd basically already have to have DNA sequences from that person or a close relative, as well as your own. And that's a pretty tall order!

Non-recombinant DNA

With non-recombinant DNA, the recombination step mentioned above never happens: what this means is that non-recombinant DNA passes from parent to child with few or no changes. Since it changes so little, our non-recombinant DNA can be very similar to that of quite distant ancestors. The key is just which ancestors.

There are kinds of non-recombinant DNA: Y-chromosome DNA and mitochondrial DNA.

Y-chromosome DNA (Y-DNA for short) is carried only by men. I got mine from my birth father, who got his from his father, etc. So Y-DNA follows the male line (see this illustration for clarification) and any two men who are related on the paternal line would have identical or near-identical Y-DNA. Men who are related but not related through the paternal line (e.g. a man and his sister's son) shouldn't expect to have similar Y-DNA.

Mitochondrial DNA (mtDNA for short) is carried by everyone. It is the DNA of tiny things called mitochondria that sit inside all our cells, and are all ultimately copies of mitochondria in the egg cell from our mother. I got my mitochondrial DNA from my birth mother, who hers from her mothers, etc. So mtDNA follows the female line (see this illustration for clarification). Any two people who are related on the female line would have identical or near-identical mtDNA.

See here for a combined picture of the female and male lines.

Using non-recombinant DNA for adoption search

Well, this is all fine and everything, but what use is it to the searching adoptee? If Y-DNA and mtDNA is the same between generations, a lot of people must have almost the same Y-DNA and mtDNA. So what use is it for finding anyone?

The answer for mtDNA is, unfortunately, not much. It can be very interesting for finding out what female-line ancestor was up to a few thousand years ago, but it's hard to use it to find someone today.

However, Y-DNA is different, not for a scientific reason but for a cultural one. In Western culture we get our surname from our father, so surnames are passed along the male line, the same way Y-DNA is inherited.

Obviously this correspondence isn't perfect—any adoptee is well aware that you might inherit a surname from someone you're not genetically related to—but it does hold to some extent. It doesn't work the other way: not all Smiths stem from one original Mr. Smith, so not all Smith men will have the same Y-DNA.

There are databases of people you have paid or volunteered to have their DNA sequenced, and many of the companies that offer sequencing services will give customers a list of close matches, with surnames included.

Remember that Y-DNA doesn't change much between generations. If you're a male adoptee of European ancestry and there's some man in the genetics database who is a male-line relative of your birth father, there's a good chance he will have the same surname as your birth father. And it doesn't need to be a close relative: as long as your common male-line ancestor lived sometime after surnames started to be used (about 1500 or so) he will likely have the same name.

So—hopefully you're still with me—the conclusion is that genetics can be a useful way for male adoptees to find the surname of their birth father. But you have to be lucky.

For accounts of adoptees or people in similar situations actually doing this, here's a 2005 story of a donor-conceived American boy finding his genetic father through Y-DNA, and here's a 2006 BBC story about using Y-DNA to predict a surname for a criminal investigation.

Rather more recently, here's the story of a Michigan adoptee named Richard Hill who used Y-DNA to find his birth father. There is also a promotional video featuring Hill produced by Family Tree DNA, the company who sequenced his DNA.