![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
One of my most vivid memories is of a rest break in the middle of the Russian proficiency test which a dozen of us graduate students in the Slavic Department were taking to fulfill the language requirement for a masters degree. We all stood up to stretch our legs. I saw one of the women, who I admit I was very fond of, glance oddly around the room and then exclaim in a loud surprised voice to no one in particular, "I'm the only one in the room, who doesn't have blue eyes!" It seemed to be an odd thing to be interested in at that moment. I looked around quickly and saw a lot of worried looking young men and women, who even after the outburst were not paying much attention to eye color or anything but their own thoughts. The speaker obviously wasn't very worried about the section of the test we'd just taken, concentrating on everyone's eye color, and I guess I wasn't either, concentrating on her and her deep brown eyes.
One of the first things we learned about genetics in primary school was that blue eyes are inherited. We were told that you had to have two genes for blue eyes and thus both of a blue-eyed person's parents had to carry one gene for that trait. It made perfect sense in terms of my family. My father, my sister and brother, and I all had blue eyes. But my mother had hazel eyes, speckled and streaked with lots of different colors, blue, yellow, green, rusty brown. My father had two of the genes for blue eyes, and by chance all of us children inherited my mother's gene for blue eyes. No need for 'further study' as they say. Several years later I was in high school and trying to get the attention of a different young lady. I gave her something I'd written that included something I thought was a nice compliment about her brown eyes. Her first reaction? "I don't have brown eyes! I have hazel eyes!' Oops! In my defense, her eyes were indeed brown in color, a nice milk-chocolaty brown, not the deep brown most people think of with brown eyes. And hazel? Well, if she and my mother stood side by side, no one would ever say their eyes were at all the same color! So my estimation of what hazel eyes should look like and the girl's hazel eyes were wildly different. I was too embarrassed to be thinking about genetics at that moment, but clearly I should have noticed before then there is a lot more to eye color than blue eyes on, or blue eyes off.
Since my grade school days, science has learned that individual genes aren't the end all and be all of inherited traits. But that is so complicated that comparatively little of that has been discovered yet.
Which brings us to the subject of the popular DNA tests offered by a number of companies for a non-trivial fee. There is plenty of useful information one of these tests could give you: finding unknown relatives, telling you about risks for genetically transmitted diseases and so on. But one of the biggest selling points the companies give is mapping out your genetic origins: where 'your people' came from. This kind of analysis is a considerably less accurate than they would lead you to believe.
First of all we would *like* to believe that the populations from various places on earth are so distinct and separate that it is an easy task to tell ancestry from here from ancestry from there. (Especially if you happen to be a big time racist!) The simple fact is that statistically speaking compared to a lot of other mammal species there isn't a heck of a lot of difference between any humans genetically, African, native Australian, East Asian, European, or thorough mixture. So what the companies are working with is a very limited number of genes and gene sequences that seem to be characteristic of particular geographic locations. Now mitochondrial DNA can tell you very precisely which woman your mother's mother's mother's mother's mother's ad infinitum mother came from. Very interesting. But that is only an extremely tiny portion of your family history. If you are male, your Y chromosome can tell you similar things about you father's father's forefathers, but again only one direct line and if you are female (and don't have a close male blood relative handy who can be tested) you are out of luck on that score. Assuming an average of 25 years between generations, five hundred years ago each of us potentially had over a million individual ancestors. Compared to that, one or two easy-to-define direct hereditary lines are almost trivial.
So let's talk percentages. When you take a DNA test the company will tell you what percentage of your ancestors came from areas of the world x, y and z. This a not exactly a lie, but it is misleading. The one fact that is drummed into our heads when learning the absolute basics of genetics is that we inherit (y chromosomes with their direct X pairs excluded) exactly half our genes from our mothers and half from our fathers. No question, this is basically true. But in our minds we tend to think that by extension we must get exactly 1/4 of our genes from each grandparent, 1/8 from each great grand parent and so on. This is a very big mistake. From each parent we get half their genes from the pool of genes they got from their parents. Mathematically speaking, whether you are male or female, it is possible for you to inherit 100% of your mother's genes from her mother and none uniquely from her father. Statistically speaking that's very close to impossible, but technically it's not impossible. What you get from each parent is a random selection from their internal pool of genes, not a definable amount from each of your more distant ancestors. Since the vast majority of our genes as humans are so similar it generally isn't a big deal. But when you are looking at a much smaller selection of genes for this or that, it becomes much more important. Imagine that we are going to look at 100 genes from our personal DNA. We look back at our mother's parents' genes and can say that statistics show that most of us, around 68% of us, will get somewhere between 32 and 68 of these 100 genes from each. Around 26% of us will get between 5 and 32 from one parent and 68 to 95 from the other. Finally over 5%, one in 20 people, will get 5 or less from one of mother's parents and 95 or more from the other. So it is very possible that one of your parents' parent's gene sets (particularly in our small sample of genes) will show up in you much more prominently than each of those from the other three individuals!
What does this mean for those DNA test results? It means that they can only give you data on what you actually inherited, not on what your ancestors' DNA looked like. So suppose grandma told you a story that one of grandpa's grandmothers was a full blooded Choctaw Native American. Claiming Native American ancestry is more common than actually having Native American ancestry and having it is very difficult to prove through documentation. If your DNA test could be perfect and grandma was right, the test would hopefully tell you that you were at least 5% Native American by descent. But the number of genes they are testing for is small, and presumably in this case, the number of great-great grand parents who were Native American is one. So there is a fair chance that the test results will show 0% Native American ancestry even if grandma knew exactly what she was talking about!
If you want to take the test please do. It may tell you things you never knew. But don't take the ancestry results too literally. The tests can tell you a lot about who you are, but not as much about where you came from as you might hope.
One of the first things we learned about genetics in primary school was that blue eyes are inherited. We were told that you had to have two genes for blue eyes and thus both of a blue-eyed person's parents had to carry one gene for that trait. It made perfect sense in terms of my family. My father, my sister and brother, and I all had blue eyes. But my mother had hazel eyes, speckled and streaked with lots of different colors, blue, yellow, green, rusty brown. My father had two of the genes for blue eyes, and by chance all of us children inherited my mother's gene for blue eyes. No need for 'further study' as they say. Several years later I was in high school and trying to get the attention of a different young lady. I gave her something I'd written that included something I thought was a nice compliment about her brown eyes. Her first reaction? "I don't have brown eyes! I have hazel eyes!' Oops! In my defense, her eyes were indeed brown in color, a nice milk-chocolaty brown, not the deep brown most people think of with brown eyes. And hazel? Well, if she and my mother stood side by side, no one would ever say their eyes were at all the same color! So my estimation of what hazel eyes should look like and the girl's hazel eyes were wildly different. I was too embarrassed to be thinking about genetics at that moment, but clearly I should have noticed before then there is a lot more to eye color than blue eyes on, or blue eyes off.
Since my grade school days, science has learned that individual genes aren't the end all and be all of inherited traits. But that is so complicated that comparatively little of that has been discovered yet.
Which brings us to the subject of the popular DNA tests offered by a number of companies for a non-trivial fee. There is plenty of useful information one of these tests could give you: finding unknown relatives, telling you about risks for genetically transmitted diseases and so on. But one of the biggest selling points the companies give is mapping out your genetic origins: where 'your people' came from. This kind of analysis is a considerably less accurate than they would lead you to believe.
First of all we would *like* to believe that the populations from various places on earth are so distinct and separate that it is an easy task to tell ancestry from here from ancestry from there. (Especially if you happen to be a big time racist!) The simple fact is that statistically speaking compared to a lot of other mammal species there isn't a heck of a lot of difference between any humans genetically, African, native Australian, East Asian, European, or thorough mixture. So what the companies are working with is a very limited number of genes and gene sequences that seem to be characteristic of particular geographic locations. Now mitochondrial DNA can tell you very precisely which woman your mother's mother's mother's mother's mother's ad infinitum mother came from. Very interesting. But that is only an extremely tiny portion of your family history. If you are male, your Y chromosome can tell you similar things about you father's father's forefathers, but again only one direct line and if you are female (and don't have a close male blood relative handy who can be tested) you are out of luck on that score. Assuming an average of 25 years between generations, five hundred years ago each of us potentially had over a million individual ancestors. Compared to that, one or two easy-to-define direct hereditary lines are almost trivial.
So let's talk percentages. When you take a DNA test the company will tell you what percentage of your ancestors came from areas of the world x, y and z. This a not exactly a lie, but it is misleading. The one fact that is drummed into our heads when learning the absolute basics of genetics is that we inherit (y chromosomes with their direct X pairs excluded) exactly half our genes from our mothers and half from our fathers. No question, this is basically true. But in our minds we tend to think that by extension we must get exactly 1/4 of our genes from each grandparent, 1/8 from each great grand parent and so on. This is a very big mistake. From each parent we get half their genes from the pool of genes they got from their parents. Mathematically speaking, whether you are male or female, it is possible for you to inherit 100% of your mother's genes from her mother and none uniquely from her father. Statistically speaking that's very close to impossible, but technically it's not impossible. What you get from each parent is a random selection from their internal pool of genes, not a definable amount from each of your more distant ancestors. Since the vast majority of our genes as humans are so similar it generally isn't a big deal. But when you are looking at a much smaller selection of genes for this or that, it becomes much more important. Imagine that we are going to look at 100 genes from our personal DNA. We look back at our mother's parents' genes and can say that statistics show that most of us, around 68% of us, will get somewhere between 32 and 68 of these 100 genes from each. Around 26% of us will get between 5 and 32 from one parent and 68 to 95 from the other. Finally over 5%, one in 20 people, will get 5 or less from one of mother's parents and 95 or more from the other. So it is very possible that one of your parents' parent's gene sets (particularly in our small sample of genes) will show up in you much more prominently than each of those from the other three individuals!
What does this mean for those DNA test results? It means that they can only give you data on what you actually inherited, not on what your ancestors' DNA looked like. So suppose grandma told you a story that one of grandpa's grandmothers was a full blooded Choctaw Native American. Claiming Native American ancestry is more common than actually having Native American ancestry and having it is very difficult to prove through documentation. If your DNA test could be perfect and grandma was right, the test would hopefully tell you that you were at least 5% Native American by descent. But the number of genes they are testing for is small, and presumably in this case, the number of great-great grand parents who were Native American is one. So there is a fair chance that the test results will show 0% Native American ancestry even if grandma knew exactly what she was talking about!
If you want to take the test please do. It may tell you things you never knew. But don't take the ancestry results too literally. The tests can tell you a lot about who you are, but not as much about where you came from as you might hope.