Usually when two genes demonstrate different topologies, there is usual explanation that genealogies of these two or more loci were different, they may split in different time, be a subject of selection, introgression etc. I have different trees build on different mitochondrial coding genes (Cytb and COI) and wonder if there was any publication discussing why it may happen. Since both genes are on mt genome, this is one locus, so introgression and differences in splitting time could be ruled out.
I suppose selection would be one possibility, but it seems odd for these genes. Before reaching that conclusion, however, I would suggest carefully excluding several other potential factors (I guess you have probably looked at these already, but they might be of interest to others):
1. Are the topologies *significantly* different? i.e., are there nodes that are strongly supported in one tree that are strongly opposed by the other? Some differences are inevitable as statistical "noise", and if all you are seeing are differences in nodes with low branch support, then that is likely to be your explanation.
2. Any possibility of nuclear pseudogenes? They certainly can happen, and they aren't always obvious. You could try using different PCR primers for the same genes to see if you get the same result.
3. Lab screw-up (sample labelling)? Happens to all of us, depending who did what. I have no idea if that is plausible for your dataset.
It's obviously very difficult to know if any of these are at all likely without knowing anything about your dataset, the nature of the differences, etc., but perhaps they are something to think about.
Dear Wolfgang, thank you for your answer. I can almost certainly exclude 2 and 3 - I had multiple samples to be sure that there is no mislabeling, no indirect evidences of NUMTs and I know how it looks when you have nuclear copy. Well, bootsrap support values for the nodes of "fast and dirty" ML tree are pretty low, you might expect such incongruence. But can I independently check for different selection pressure in these genes?
Dear Artur, trees are attached.
- Badges
- Science method