Page 1 of 1

measuring genetic diversity

Posted: Sat Jan 30, 2010 2:00 pm
by deleted-29930
For my project I am modeling how human genetic diversity varies with distance from Africa. All articles that measure human diversity talk about
mean microsatellite variance. Could someone explain what this is, how it is measured and what statements like R^2 and P mean. Thank you

Re: measuring genetic diversity

Posted: Sat Jan 30, 2010 2:32 pm
by deleted-71536
Hi there,

What an interesting project! It's true that the articles about human diversity use a lot of scientific jargon, and it can be confusing. Let me break down the terms for you.

Mean = This is the scientific word for "average," calculated by adding your figures and dividing by the number of samples.

Microsatellite = Microsatellites are regions in DNA where the code repeats. Because of these repeats, sometimes there are mistakes in DNA replication that can either lengthen or shorten the number of repeats in that area of the DNA. (You can think of it like this: with so many little repeats, the DNA polymerase molecule can get confused and not remember how many repeats to make.) Due to the tendency for these regions to be lengthened or shortened during DNA replication, they are highly variable and mutate often. This means that they can be used to look at genetic differences in closely related organisms. Here is the Wikipedia article on microsatellites to get you started: http://en.wikipedia.org/wiki/Microsatellite

Variance = This is a scientific way to measure how variable the data are, or how far away most of your samples are from the mean (average). High variance means that many samples measure far away from the average, while a low variance means that most samples are close to the average.

R^2 (r squared) is a statistical measure of how well your data are correlated. If the data fit a regression very tightly, you will get a higher R^2 value. If the data are very scattered, the R^2 value will be lower. Here is the Wikipedia article on R^2: http://en.wikipedia.org/wiki/Coefficien ... ermination

A P value is a statistical measure of how likely it is that your results were due to chance. By convention, scientists usually consider results to be statistically significant if the P value is less than 0.05. This means there is less than a 5% chance that the results found were due to chance alone. Here is the Wikipedia article about P values: http://en.wikipedia.org/wiki/P_value

You may have already found this, but here is another article about using microsatellites to look at human diversity:
http://www.pnas.org/content/94/7/3100.full

I hope this helps. Please post again (in this same topic) if you have more questions about your fascinating project!

Heather