Parsimony score of a NJ tree

Question

[mdsiddra]

I have been working in evolutionary genomics field and hence reconstructing the phylogenetic trees using nucleotide/protein sequence data. I need to ask can a calculation of parsimony score be done for a Neighbor-joining or UPGMA tree to find that topolgy of the tree which represents the minimum evolutionary changes???

Answer 1

[grant.hovhannisyan]

NJ and UPGMA are distance matrix-based methods. So if you only have the resulting tree obtained by these methods and don't have the sequences, than you can not perform parsimony calculations.

[mdsiddra]

Lets say we have a particular set of sequences, and a NJ / UPGMA tree for this data too. Can this tree be used as an initial tree topology for parsimony score calculation to result a tree that has minimum evolutionary changes?

[grant.hovhannisyan]

I am not aware of any method/software for doing that, but the problem I see is that the NJ tree topology might not coincide with the real maximum parsimony tree. Why not just to reconstruct the parsimony tree if you have sequences?

[mdsiddra]

As small parsimony problem states that it works with 'n' number of sequences and an initial tree topology and then uses the heuristic searches (NNI, SPR, etc) and calculates the parsimony score resulting the tree with an arrangement of least score. Moreover, as suggested that Neighbor joining is just a clustering algorithm that clusters haplotypes based on genetic distance, using this as basis, In this scenario can a tree topology reconstructed using NJ method be used as an initial tree topology for MP method?

[grant.hovhannisyan]

Yes I know what you mean, but I don't know if this option exists in MP software. You can have a look in the documentation of the software that you are using (I guess PAUP).