Is there a good reason for multi-locus phylogeny when there is PHT conflict?

Question

[Yongjie Zhang]

Hi all,

I am performing phylogenetic analyses using 7 different fragments for the same set of individuals (all same species). By comparing the single-locus tree visually, no any two trees have identical topologies. PHT test by PAUP showed no significant difference among the first 6 fragments, but significant difference (P=0.001) between the 7th fragment and each of the first 6 fragments. I prefer to use the combined data set of all 7 fragments because all my individuals are belong to the same fungal species. But does anyone know a good reason for adding the 7th fragment (I need a reason when writing my paper), or any other test I can do? Thanks.

[Joseph Hughes]

I think the fact that you have found a gene to be significantly different to the other six should not be glossed over, there could be an important biological reason for this such as recombination or horizontal gene transfer. I think that these results should be presented in your paper. Subsequently, you can decide to create a super matrix combining all your genes stating that the issue of combining data for phylogenetic reconstruction has still not reached a consensus with some recent references of the ongoing debate.

[Yongjie Zhang]

Joseph, thanks for your reply.