Here we describe some situations when phylogenetic tree reconstruction is needed.

Some psychologists claim, there is a functional relationship in males between IQ score and brain size, and these are in negative relation with level of testosterone, or weight of the testicles. This concept arose from comparing such variables among human races. Can we simply calculate correlation on such a data set?

Some biologists claim, that flexibility of behaviour is related to the size of association areas in birds or mammals. Taking measures of brain sizes and appropriate scores for behaviour flexibility for a bunch of species can we simply calculate correlation among such variables?

Independent comparison

Assume you need to find out whether two variables, A and B were correlated by comparing four species, 1, 2, 3, and 4. The measures for the two variables are here in the table.
By plotting these data points we have the impression, that A and B are significantly correlated to each other.

Phylogenetic tree reconstruction

To see why the data points in the graph are not independent, simply move your mouse over the figure. The points represent species, species sit on a phylogenetic tree, therefore they are related to each other to some degree. To calculate correlation, indeed, to do almost any statistics on such data we have to get rid of the effects of common descent. For that, however, we need to know the structure of the tree.

Origins

Besides comparing species, there are many other cases when we need a phylogenetic tree. Sometimes we are simply curious "who" is our closest relative, or we may want to know which living species is the closest relative of birds.

Or consider the case of the Florida dentist, who might or might not have infected his clients with AIDS virus. Virus samples were taken from the dentist, the clients and other locals, and the DNA was sequenced. As viral DNA mutates fast, it is not easy to show which variant could be the ancestor of all the other DNA. For that a phylogenetic tree had to be constructed. The result is not without controversy, which is not surprising in the field of molecular trees (see one argument on the web)

The methods to construct phylogenetic trees are abundant. In this course we focus on molecular trees constructed from DNA sequences.

Limitations:

There are a number of theoretical and practical problems with DNA trees. First of all, you should bear in mind, that trees constructed on a limited number of DNA sequences reflect the evolution of that particular gene, and not necessarily the phylogeny of the taxa. Trees constructed on the basis of a single sequence are therefore called "gene trees".

Secondly, the rate of evolution might differ among taxa, among genes, and gene regions. There are methods to estimate and control for such discrepancies, and in real life one always should check for such effects.

Practical problems include mistakes in the sequences, limited knowledge about the sequence, shortcuts in the analysing methods because of excessive computer time etc.

The steps:

The main steps are: 1. finding genes 2. aligning them 3. constructing tree on the base of an evolutionary model. Than you can do evolutionary independent comparisons.

Each section starts with a short introduction to the problem, than you can go on to a help page and navigate through databases, software. You get concrete problems with solution, and than practising exercises to be solved on your own.

Background:

The Workshop on Molecular Evolution at Marine Biological Laboratory provides a great Glossary of terms. The Glossary was built on the book by Li & Graur. As we found this book extremely useful, we also recommend the glossary. Links to more glossaries are listed in our Link Collection.

GO TO: Databanks

Salve evolution links
Salve bioinfo links
Salve evolution news
Salve genetics news