Phylogenetic Tree Read And Interprettion

Question

[Oumed]

I appreciate to help me how to interpretate the phylogenetic tree after it is built?

what does the numbers on the nodes mean?

How to know the % of similarity of two sequences in the same clade?

Best regards

oumed

Answer 1

[Niek De Klein]

You should interpret the tree by using all the possible information that you have/can get. Knowledge beforehand about the organisms in your tree is a very big plus. If you have that knowledge you can judge if what your tree is showing makes sense.

Some important things to look at: Is the outgroup actually on the outside of the tree? Do the placing of the organisms make sense (if you have a bacteria in a branch under eukaryotes there might something be wrong). If you have DNA or protein sequences from gene families, are the gene families grouped together?

These are some very general advices, you should judge a tree differently depending on the sequences you use to make it. You would look at a tree differently with very close related sequences than you would with sequences that are far apart.

---

I find the best way to know the % similarity is to, if you have that information, look at the multiple sequence alignment. You can look at the distance between two sequences(I suppose that that's what the numbers on the nodes are, although they usually are on the branches where they split off) but the alignment holds more information than just the distance score.

The number on the nodes are the distance scores I presume. I expect them to be where the branches split off, but I guess that might depend on which tree viewer you are using. I haven't worked with phylogenetic trees in a while so I'm not 100% sure (but you can look that up) the number shows the % of difference between two sequences (although I think this might differ between different tree viewers, again, not sure). A score of 0.01 means that there is a difference of 1% between two sequences.

---

I'm a bit rusty on the phylogenetic trees, but there is loads and loads of information about them. A bit heavy reading but a thorough explanation of the scoring: http://mbe.library.arizona.edu/data/1994/1102/11taji.pdf.

[Oumed]

the number on the nodes of phylogenetic tree correspondedd to bootstrap value, is bootstrap value a way of expression of similarity or distance

Answer 2

[Alex Paciorkowski]

Here is a recent phylogenetics review that I've found helpful: Sleator, 2011

Answer 3

[Niek De Klein]

@ your comment (was too long to answer in a comment myself) Oh you were talking about the bootstrap values. The bootstrap value has nothing to do with similarity or distance (ok, indirectly it has), it's a measure of how 'trustworthy' your tree is. Depending on how many bootstraps you choose (i think default is usually 100, better to do 1000) it will rebuild the tree but it will remove some sequences and rebuild it with a random sequence (not a good explanation, you should read about bootstrapping to understand it fully). If you did a 1000 bootstraps, if all values are above 900, it's a good tree. If they're all below 100, it's a bad tree (very generalizing). But there is a lot of interpretation going on when looking at the bootstrap value too.

I don't have time for a good answer now, try googling bootstrapping it's a much used statistical method, or maybe someone else can answer this in more depth.

Answer 4

[Ahdf-Lell-Kocks]

The bootstrap values signify how sure one can be that the tree is correct at any given node. A tree is calculated by looking at existing molecular data, and the more data one can gather, the easier it will be to relate the sequences in a highly-confident phylogeny. If the sequence evidence is not good enough, the bootstrap values will be low.