Pairwise distance of rooted phylogenetic tree

Question

[Chirag Parsania]

Hi,

I am quite new to phylogenetic analysis. Please bare with me if question does not make sense.

I am using IQtree to analyse some horizontally transferred genes. I have successfully generated phylogenetic tree using IQtree. I have few questions related to interpretation of the trees given by IQtree.

1). I want to find relatedness between each pair of species in the tree. In simple terms, I want pair wise distance matrix. To do this, first, I converted unrooted tree to rooted tree using outgroup sequence. Question is, how can I get the pairwise distance of rooted phylogenetic tree ?? I know IQtree gives .mldist file having pairwise distance matrix of final tree given by IQtree. However, I don't know how much the matrix of .mldist file is reliable to use if I converted original unrooted tree to rooted tree ??

Any answer would be highly highly appreciated.

[Joe]

Phylogenetics on horizontally transferred genes is a whole other ball game.

Are you trying to infer phylogenetic relationships about the parent sequences, or just the similarities/differences of these genes specifically?

Check out the ETE toolkit for manipulations/analyses of trees in any case.

[Chirag Parsania]

Hi thanks for reply. I am trying to compare species tree vs gene tree. However, I have 100s of horizontally transferred (HT) genes an therefore it is not practical to make pair of species tree and gene tree for each of them and then compare with gene tree. To address this, I designed alternate analysis. Please correct me if there is any flow.

Rather making 100s of tree, I though to make a scatter plot of gene distance vs species distance. Scatter plot includes two types of orthologoues gene sets. 1)HT genes and 2) as a control, I choose genes which are evolutionary conserved genes; you can say non horizontally transferred genes. With these two gene sets, what I assume that, for the evolutionary conserved genes, scatter plot of gene distance vs species distance will show very high correlation. However, for the HT genes species distance will remain same but gene distance will be very less.

So the question here is which distance should I choose as gene distance from the following ?

1) Take the pairwise distance matrix given by the IQ-tree programme 2) Convert un-rooted phylogenetic tree to rooted phylogenetic tree and then calculate the distance of most recent common ancestor (MRCA) 3) Convert rooted phylogenetic tree to distance matrix using cophenetic R function and then use that distance matrix as gene distance

Sorry for very long reply.

[Joe]

I'm not sure I totally followed this, but if my understanding is correct, that seems a reasonable approach. It sounds a little like a dN/dS analysis to me? You're supposing the the HGT-derived genes will not correlate as well with the 'species distance' (not sure what you mean by species distance though)?

My gut feeling would suggest approach 3, but it's been a while since I've looked at cophenetic correlations so I may be mistaken on this approach.

If you are simply interested in the pairwise distances amongst your orthologues, I would be inclined to think rooted versus unrooted doesn't actually matter in this case, unless I'm missing something. Inclusion of your outgroup would presumably appear as an outlier in the similarity matrix depending on exactly how far away it is?