Question

37| Sample3 44 Sample 71 Sample4 Sample5 Sample 1 Sample2 0.10 Fig. 1. Neighbor joining tree for Bacillus thuringiensis samples 59 Sample2 32 Sample5 Sample 1 Sample4 Sample3 96 Sample6 Fig. 2. Maximum parsimony tree for Bacillus thuringiensis samples

Hello! May I know how to interpret the data from this figure? I'm having a hard time understanding it and figuring out the evolutionary relationships between the branches.

Answer 1

Okay, let us analyze how a phylogenetic tree works. The branch tips are always going to indicate the analyzed groups, in this case they are samples. When 2 samples are the closests between them they are going to have only one node between them (nodes are where branches become one single branch) and they are going to be called "sister" samples. The more nodes you find separating 2 samples the less close they are.

Now, parsimony method applies the rule of simplicity, that is to seek the most simple tree to explained the relationships according to the data, the tree that has the lowest number of steps.

Neighbor joining methods don't use the simplicity rule, what it does is to use distance matrices (based on a character matrix) in which the lower distances will determine the sister species/samples.

Now, the number in the nodes (note that they ARE associated to a node, always a group of samples) is the group support. A value of 100 would indicate that this group has a support of 100%, that means that with this data and method we can be sure the group exists. As the value goes down from 100 we can see different probabilities for the groups for really being reliable. For example, in your Neighborjoining tree, the groups containing samples 3, 6, 4 and 5 has a somehow high support of 71, while the group that includes only 3 and 6 has a low support of 37 which is not realiable because it is too low.

But when you analyze the parsimony tree you can find that this method actually better supports the relation between 3 and 6, with a value of 96, which is already significant.

I hope this is of help for you to analyze the tree. Remember that the trees are only hypothesis of relation, that follow only the method you used to build them and the data you used, but the results may vary if you use other method or include different data.