Wednesday, April 02, 2003

There is possibly nothing more fascinating that the modern study of genetics

If you look at certain sections of the DNA human beings, and then record the variation in the frequencies of particular gene patterns over wide areas of Eurasia, it is possible to see clear variations depending on where you are, but it is hard to find patterns. However, if you use a mathematical technique called principal components analysis (PCA) to filter the data, patterns become much clearer. Loosely, what PCA does is attempt to take the complex patterns on the initial map, and figure out what is the simplest way to generate the complex patterns as the sum of five or six simple patters.

Once the PCA analysis done, you are left with is a series of five or six contour maps, each representing one of the simple patterns that can be added together the generate the complex pattern.

This work was initially done by the Stanford based team of Italian geneticist Luigi Luca Cavalli-Sforza. Having got these maps, the results were quite extraordinary. One of them was found to correspond very closely to the arrival of cereals in Europe as shown by radiocarbon dating. Another spread out through Europe from an area north of the Caucasus and the Black and Caspian Seas, and appears to show the origins of whoever brought the Indo-European language family to Europe. Yet another is centered on the Basque country of northern Spain and southern France, which is very interesting given that this area is unique in genetic and linguistic terms.

It seems clear that the principal components of the genetic data show the invasions of Europe by various groups of people over the last ten to twenty thousand years. That is right: by looking at people's genes, you can figure out a great deal about just how mankind spread throughout the world.

Looking at Asia rather than Europe, similar analysis shows for instance that Pacific Island people originated in Taiwan. Very reassuringly, linguistic evidence from the analysis of various human languages shows much the same thing. Similar analysis shows the spread of various ethnic groups throughout Africa as well.

This is fascinating stuff, and is really remarkable. Cavalli-Sforza covers it all quite well in his book Genes, Peoples, and Languages, which I thoroughly recommend. The book also contains the actual pictures I have described here but have not been able to show. I would scan a couple of them, but my copy of the book is (unfortunately) in Australia.

This article from the Guardian by Professor Johnjoe McFadden of the University of Surrey (via aldaily attempts to summarise some of the results that have been derived from this type of analysis, but in my mind does not do so very successfully. It has a strange, rather clipped quality, and it manages to give us a pot pourri of results but goes through its argument so fast that one almost feels the author initially wrote something twice the length and then had it chopped in half by his editor.

Plus, it makes one very misleading statement at the start. Specifically


Rebecca Cann, of the University of California at Berkeley, was one of the first to use DNA to uncover the past. In 1987 she examined the mitochondrial genes we inherit from our mothers and showed that our female line can be traced back to a single woman who lived in Africa about 200,000 years ago, a "mitochondrial Eve". Fossil evidence for modern humans goes back at least 500,000 years so it seems that humanity went through a severe bottleneck when this Eve was alive - it remains a mystery why of all the females who inhabited the globe at that time, she was the only one to leave modern descendants.


Mitochondria is passed from the mother to the child. Therefore, you have the same mitochondria as your mother, your maternal grandmother, your maternal grandmother's mother et cetera, going up the female line. Whereas most DNA comes equally from the mother and the father, and varies from parents to children through sexual combination, mitochondria is simply passed from mother to child. The only changes are due to mutation. Because of this, mitochondria changes more slowly over time than other human DNA, and it is therefore of particular interest to geneticists

If the mitochondria of everyone in the world is studied, it can fairly quickly be shown that the mitochondria of everyone in the world has a common ancestor about 200000 years ago. What this means is that if you go back along the female line from everyone in the world to this time 200000 years ago, then you always get to the same person. Everyone in the world is descended along the female line from this one person. When Rebecca Cann, of the University of California at Berkeley, made this discovery in 1987, this ancestor was nicknamed "Eve", which is unfortunate, because the biblical Eve was supposedly the only woman in the world. The mitochondrial Eve was not like this. At the time she lived, there were lots of women in the world, and every person in the world today is descended from almost all of them. However, not down a direct female line. The paths by which you and I are descended from all the other women in the world at that time are not strictly female lines, but involve a mixture of men and women along the way. Rather then being descended via your mother's mother's mother, you might be descended by your mother's father's mother. It is not true that "of all the females who inhabited the globe at that time, she was the only one to leave modern descendants". It is only true that she was the only one to leave modern descendents by a direct female line. (I again tend to think an editor of the Guardian might have shortened what was originally written without really understanding it).

And, if you look at the mathematics of breeding, then this is not surprising. If you look statistical patterns of people mating over hundreds of thousands of years, you find that in fact it is certain that all the female lines will intersect at some point. Dr Cann's research was impressive, but what she did was pin down when the female lines diverged. Everybody knew already that it happened at some point.

Update: Coincidentally, today is the fiftieth anniversary of Watson and Crick's paper on the structure of DNA. The amount that has been achieved in the 50 years since its publication is simply mindblowing. (Link via slashdot).

No comments:

Blog Archive