Starting With the Answer (Becoming a DNA time traveler...)
Typical genealogy research involves pushing back your ancestry lines, generation by generation, to discover your ancestors. How would it change your research approach if you started with the answer - if you knew who from the early 1700's you descended from, for example, but didn't know the details of how. If that sounds like fun to you and you have $150-$300 burning a hole in your pocket, read on...
Researching my genealogy is a hobby I've always done for fun. I like history and there's something about understanding my personal connections to it that makes it even more interesting. I make my living from technology, so when I heard that DNA analysis could be applied to Genealogy... I was curious on a couple of levels, and decided to get tested. The following essay covers what I've learned from the experience. Speaking as a lay person, I'm no doubt over-simplifying and perhaps even unknowingly fudging a fact here or there. As with all information... consider the source.
DNA are the bits of biological information we all carry around that makes us unique - the stuff that determines, out of all the possibilities our bodies have... which specific set of characteristics an individual is given to work with. You get roughly half of this stuff from your mom and half this stuff from your dad, and in turn contribute half of what you got to your children. Which half? Well, nature loves both trial and error and continuing a good thing, so some portions are passed randomly, some portions are passed strictly father to son, mother to son, and mother to son and daughter. It's these "strict" portions that we're interested in for genealogy research purposes, because they prove a relationship.
The amount of information contained in your DNA is huge... even small portions of it contain enough information to uniquely identify you apart from any other person. That's what a paternity test is - an analysis of a portion of the stuff that's passed parent to child. Now for the big ah-ha... if this stuff is unique between a father and a son, wouldn't it be unique for a grand-father and a grandson? YES. How about a great-grand-father and a great-grandson? YES. Now, before you get too excited, observe that this method of identification only works for one of two grand-fathers I have. Why? Because the DNA my father and I share obviously isn't shared by my mother's dad (they're not related biologically at all, or at least I hope they weren't :-)... so it requires an unbroken chain of the same-sex to work for identity purposes- either male to male to male or female to female to female.
Ok, if you're a deep thinker, at this point you might be wondering... "Hey, if a human male always passes these same bits to his son, why wouldn't all human males have the same bits?" Well, it turns out that every so often, nature, who loves trial and error, randomly tweaks these "strictly passed" bits so they're ever so slightly different between father and son (or mother and daughter - I'll explain later why I'm use father-son examples). Over a long time span, these occasional tweaks add up to some vary diverse bits of information.
This occasional diversity is a boon for genealogists... since this rate of change is rather predictable and steady. In the commonly used male to male DNA bits, the rate of change is generally one change every 7 or 8 generations. So if a fellow researcher of my surname and I share the exact same bits, then we know we likely share a common father in the past 7 or 8 generations. If our bits are one step different, then it's probably 7-15 generations ago that we had a common father, etc. Different bits of the DNA do vary at different rates, so you can get a bit more precise by looking at the specific bit that's different.
Ok, back to you deep thinkers... "Wait, if we know the rate of change of these bits of DNA, couldn't we do a sampling of all current DNA in the world and do some fancy math to figure out how long it took for it to diverge this much? And wouldn't that say something about the origins of all of us?" Yes, in theory. The Seven Daughters of Eve by Bryan Sykes is seen as one of the seminal books on that subject for lay people, if you're interested.
Getting back to genealogy... my mom's a Fike. She calls herself a Fike because that was her name growing up - her identity. She looked like and acted like people of that name who lived in the area because they were all part of the same extended family. For DNA identity purposes, my mother's not a FIKE. Why? Remember DNA identity only works along strictly male or strictly female lines. FIKE is a surname. Surnames are passed down from male to male to male. So while my mother has a unique DNA that ties here to her mother (a DAVIS) and to her mother before her (an ENNIS)... most cultures don't assign a consistent name to that. To make things tougher for female research, the strictly passed female to female bits of DNA don't vary much over time (compared to the male to male bits) so it's harder to pin down the time frame of a relationship. So most people using DNA for genealogy purposes are doing so for surname research, which is male to male to male... the same as those unique "male to male" bits of DNA. That's why I used father to son examples earlier.
People of the same surname form "surname projects", typically to try and collect samples of that surnames DNA. With enough samples, patterns emerge as to how many different "lines" of that surname exist. For example, in the US, with a surname like Hancock, and 35 or so samples collected, a pattern showing 3 or 4 original ancestors has emerged. If you're researching a Hancock, and you can get a male Hancock from your family (or yourself if it's your surname) to test for you... chance are reasonable that you'll tie to one of those known ancestors. Now that you know your original ancestor, you can research from both ends - from the original ancestor forward, and from your known ancestor backwards. With the Underwood surname project, it took us 6 samples before we found two people that were related.
My advice, money and fun aside, is to seriously consider using DNA testing for your lines where:
a) you have a male by that surname that can be tested,
b) there's an existing "surname project" with 20 or more samples already collected,
c) and your surname isn't Smith or Jones.
Family Tree DNA is probably the most popular vendor for adhoc surname projects and you'll find their list here ( http://www.familytreedna.com/surname.asp ). Relative Genetics ( http://www.relativegenetics.com/ ) has a higher end service that's more appropriate for established groups or to support things like family reunions and such. And there are others...
You'll find there are different types of test. The male to male test is called a YDNA test, and it's offered in different depths (typically something like 12 marker, 25 marker and 37 marker). While the 12 marker test is cheaper, it's not very useful in the surname projects I've described, because 12 markers simply aren't unique enough to tell you much. I did the 25 marker test, and so far haven't had any need for more resolution than that, so would suggest you start with that. Most companies will let you "upgrade" your test later to include more markers, so you don't have to resample and retest all over again.
There are a couple of forms of the test and sampling - from scrapping your cheek with a little device that looks like a piece of hard felt to a mouth wash that you swirl and deposit in a vial. You take the sample yourself and it's easy and not painful in any way. The sample is mailed in and results come back in 4-6 weeks.
No doubt some people are worried about the privacy of all this... As you would expect, the companies that do the analysis for you are very serious about protecting it and provide a number of safeguards and guarantees. From my perspective, my DNA is already compromised - I discard it all the time in a variety of ways. If someone wanted it bad enough, they could take it without me ever knowing. So my opinion about the privacy of my own DNA is something I don't really have anyway... but thatís just my opinion.
Iím glad I had my DNA analyzed. I havenít turned up any related fellow researchers yet, but I really think itís just a matter of time before I make a connection.
About The Author: R. Aaron Underwood is a software author (the genealogy product GenSmarts Ė http://www.GenSmarts.com ) , living with his wife and three kids in Long Grove, IL, and has enjoyed Genealogy as a hobby for the past 30 years. He can be reached at aaron@GenSmarts.com