Joy's New and Hot, Week 7

Evolutionary Trees and Math

My article comes from Sciencnews.org and is about how geometry and topology are helping biologists make better use of evolutionary trees. It begins with a hypothetical situation: an oceanographer buys whale meat at a market in Japan, and is told that it is the meat of species that can be legally killed. He tests it only to determine that it is from an endangered species, which it is a crime to kill. The oceanographer goes to the authorities and is faced with this question: how sure is he that it is actually the endangered species?

Until recently, this could be a large problem, since the genome of all whales hasn't been mapped. Using evolutionary trees could be problematic because if the data provides evidence pointing towards different trees, the biologist has to determine between these which is the best. If the different trees are relatively close together, getting it wrong might not be too bad, but if the trees are far apart this could be a problem. However, Susan Holmes, a statistician here at Stanford (I always think it's cool when I read about people from here in the news), and mathematicians Louis Billera and Karen Vogtmann from Cornell have developed a new way to use evolutionary trees to determine more accurately to what species a certain creature belongs.

They made a sort of evolutionary "forest" out of the trees, creating a three dimensional representation. The biologist can take his average evolutionary tree, and trace the shortest line between it and the surrounding trees to see what his average tree is most closely related to and then pick a more accurate match for his mystery species. I'm gonna be honest, the article talks about negative curvature and some stuff I don't quite understand about the math. However, I just think it's neat to look at how different fields intersect and use one another, like math helping out biology. The researchers have used this method to create a free software program that's going to be added to a statistics program called R that biologists are already using. They are currently working on developing more accurate algorithms.