DNA Molecule

Cheek/Chick DNA Project

What It All Means

Who was our most recent common ancestor?

Scientists look at variations in DNA to study the movements of populations around the world over thousands of years  People who do genealogy, on the other hand, are usually interested in more basic questions about their own family history. The scientists talk about probabilities; we want to know the specifics.  Unfortunately, DNA testing, by itself, is not powerful enough to pinpoint any specific ancestor.  If you undertake a family tree DNA project with that goal in mind, you will end up disappointed.  It turns out that our "Most Recent Common Ancestor" is an elusive fellow.  The scientists cannot tell us exactly when or where he lived.  What they can do is give us a time frame in which he probably lived.

When the scientists talk about the Most Recent Common Ancestor, they are essentially talking about a statistical construct.  In other words, the Most Recent Common Ancestor is not a person—he is a graph, or a range of probabilities.  Okay, you are saying, that sounds pretty weird and really not at all helpful.  Therefore, stop thinking about DNA for a moment and think about dice.

When you roll a pair of dice, there are 36 different ways that the dice can land.  Only one of these is a double "1" or as gamblers say, snake eyes.  Your chance of getting a snake eyes is 1 in 36, or about 2.8%.

pair of dice showing snake eyes

Of course, even though you can calculate the odds of rolling snake eyes, you never know how the dice will actually land on any particular roll.  Every roll of the dice is random.  You could get snake eyes on the very first roll, or you could roll the dice 20 times and never get snake eyes at all.

Now, imagine that you are hanging out in a bar with your obsessive-compulsive gambler friend (we'll call him Al), and Al says to you, "I have been sitting here rolling dice over and over again, and I didn't get a snake eyes once.  So, smart guy, how many times did I roll the dice?"

Can you answer Al's question?

No, obviously you cannot, unless you were watching the whole time, or secretly videotaping Al with a hidden camera (let's assume you weren't—this isn't Vegas).  At least, you can't answer Al's question specifically.  The only thing you can do is come up with a range of probability.  In other words, you can figure out how likely it is that Al could roll the dice a given number of times and never get snake eyes.  Although the chance of snake eyes on any particular roll is always 1 in 36, if Al rolled the dice 100 times, the odds are very high (94%, in fact) that he would have rolled a snake eyes at least one time.

Rolls of Dice Probability of:
At least 1
Snake Eyes
No Snake
1 2.8% 97.2%
2 5.5% 94.5%
3 8.1% 91.9%
4 10.7% 89.3%
5 13.1% 86.9%
10 24.6% 75.4%
25 50.1% 49.9%
100 94.0% 6.0%
250 99.9% 0.1%
data from above table in a graph

If you are good at math, you can make other calculations, such as the probability of rolling exactly one snake eyes, or rolling snake eyes twice in a row, or getting any particular combination of snake eyes and other rolls.  You can also improve your estimates by using real-world evidence, such as the length of time you know Al has been sitting in the bar, how fast he is physically capable of rolling the dice, etc.

Scientists do this sort of thing when they study DNA and estimate the number of generations to the "Most Recent Common Ancestor."  Of course, DNA is much more complex than dice. But the concept is similar. Whenever living things reproduce and make new living things, they have to make copies of their DNA. The process of copying is not quite perfect.  When any given piece of DNA is copied, there is always a small chance that an error (mutation) will occur, just like with dice, there is a small chance of rolling a snake eyes.  Exactly when it will happen cannot be predicted, because it is a random process.  Every generation is another roll of the dice.  I could get snake eyes on the first roll, or the second, or the twenty-fifth—there is no way to know.  However, I do know that if I roll the dice 25 times, there is about a 50% chance that I will roll snake eyes at least once.  To put it another way, if my friend Al has been rolling dice for awhile and he hasn't gotten a snake eyes yet, there is a 50% chance that he has rolled the dice less than 25 times.  Similar mathematical calculations are used to study changes in DNA.

If my DNA is mutated, is there something wrong with me?

No!  A mutation is just a change in DNA.  Mutations are not necessarily harmful; in fact, most mutations don't do anything at all.  Occasionally a mutation may cause a disease (like hemophilia), or a condition (like deafness or dwarfism), or a harmless physical variation (like green eyes).  Sometimes a mutation may even be helpful.  For example, Europeans have a simple genetic mutation that reduces the amount of dark pigment in the upper layers of the skin.  This mutation helps children growing up in foggy, northern climates avoid vitamin D deficiency (rickets).  People with fair skin have an advantage in places like Ireland and Scandinavia, although they are prone to sun burn and skin cancer if they move closer to the equator.  Fair skin is an example of a mutation that is helpful in some environments and harmful in others.

These dramatic types of mutations, however, are not particularly useful for genealogy.  Mutations that alter a person's phsyical characteristics or cause a genetic disease tend to be either very common (if the mutation is helpful) or very rare (because everyone with the mutation dies young).  Therefore, the scientists who first pioneered the use of DNA in anthropological research concentrated on the inactive parts of the DNA molecule.  A surprisingly large percentage of our DNA is inactive or "junk DNA."  It is copied and passed along from generation to generation, but it doesn't seem to do anything.  Mutations accumulate in junk DNA at a predictable rate without causing any harm or benefit to the individuals involved.  Thus, variations in junk DNA can be used as kind of genetic clock.  Scientists aren't sure why we have so much inactive DNA, but it has been a boon to anthropologists and more recently, to genealogists.

Analyzing DNA Matches

Getting back to Y-DNA testing, assume two men with the same surname get a Y-DNA test and they match on 24 out of 25 "markers."  (DNA "markers" are pieces of DNA that frequently vary between individuals, and therefore are useful for studying family relationships.)  We suspect that these two men had a common ancestor because they have the same surname and their families appear to have come from the same geographic area.  We know that there has been one mutation—one "snake eyes"—since the Most Recent Common Ancestor.  Can we say how many generations have passed—how many times the dice have been rolled?

Trying to answer the question, "when did our Most Recent Common Ancestor live?", is like trying to answer Al's question, "How many times did I roll the dice?"  We don't have a time machine that allows us to get DNA samples from all our ancestors, so we are forced to work backwards using probability and statistics.  (Of course if we had a time machine, I guess we wouldn't need to do a DNA study at all. Never mind.)

Further complicating the problem, scientists do not know exactly how fast DNA mutates.  The average probability of mutation is apparently about 0.2%, or 1 time out of 500. However, some DNA "markers" mutate faster than others, up to 1-2% of the time.  The Most Recent Common Ancestor analysis will be different depending on the estimated mutation rate, as well as the mathematical model that is used to crunch the numbers.  So again, analyzing DNA is far more complicated than analyzing a single pair of dice.  Instead, it is like having hundreds of pairs of dice, all with varying numbers of sides!

The probability calculations shown this website (see the genetic distance tables) are generated by Family Tree DNA, using the most recent scientific undertstanding about the mutation rate of different markers.  If two men with the same surname have a perfect 25/25 match, there is a 95% probability that they are descended from a common male ancestor within the last 12 generations (about 300 years), and 99% within 20 generations (500 years).  With a 24 out of 25 match, there is about an 80% probability of a common ancestor within 12 generations (300 years) and 95% in 20 generations (500 years), but the numbers can vary slightly depending on which marker has mutated.

Coincidental Matches

When searching the Family Tree DNA database, you may find that your results closely match people with different surnames.  These matches should be viewed with caution unless there is other evidence that your families had a common male line ancestor within a genealogical time frame.  Coincidental matches, especially 12-marker matches, are not uncommon in people who are part of the same ethnic group.  In such cases, your common ancestor may have lived hundreds of years ago and will be impossible to trace.