Hello. I'm currently analyzing a plastid transcriptome produced for Karenia brevis, a fucoxanthin-pigmented dinoflagellate. My understanding of the plastid lineage for this organism, is that it contains a tertiary plastid obtained from engulfing a haptophyte. As such, plastid genes should group within or as a sister to the haptophytes, which should show a common lineage with red algae. When I looked to the literature, I found that this was the case in the studies I found that had conducted similar analyses.
Since I had very little experience with phylogenetic analysis, I chose MEGA6, as it was recommended for preliminary work, although not sufficiently configurable to produce trees for publication. After choosing a group of taxa and gathering and aligning their protein sequences for psaA, I built a ML tree with the recommended LG+G substitution model. The resulting tree is attached as .docx named "First tree".
I've also included a similar tree from a publication under the name "Published comparison" which displays a much more realistic branching pattern. It's a DNA tree for psbA, but I've found that my results don't change significantly when comparing my protein trees to DNA trees that I've produced. Also, the strange branching patterns are in all of my photosystem genes, but not rbcL. The tree I built for rbcL has a branching pattern that agrees with the plastid lineage for Karenia, ie similar to the trees in "Published comparison"
I'm frankly at a loss as to how to interpret this tree like it, and don't know why it's branching so strangely. If anyone can see what I've done wrong, or has any suggestions on what topics I should read up on to solve this, I'd be greatly appreciative of any help. I feel like I'm making a naive error somewhere, but I just can't find where.
First of all, you need to check the quality of alignment you made. Second, to compare trees you need to use at least several identical sequences. I briefly checked your trees and did not found any. Could you highlight them for non-specialist in psbA? Third, both trees have several branching points with
I don't recognize enough of your plant names, either in your tree or the published comparison tree, to understand which branches should be the same in your tree vs the published. I see Zea maize (corn) and a few others I recognize but not enough shared in your tree to published tree to spot problems. I do see that the Karenia is an "outgroup" to most species in your tree, and not so very distant in the published trees. This could possibly be due to a region being misaligned in your alignment. It does not take much misalignment to create a long branch like that.
I also note a "cloroplast) Pavlova lutheri" name in your tree which could very well cause a large problem. Symbols "(" and ")" are used in treefiles (NEXUS format for example) as part of the langauge describing how to draw the tree. If you have any characters such as ")" or "(" or ":" in your species names it usually crates a treefile that connot be read by a tree drawing program at all, you just get an error message. But it is possible that you were able to draw a tree that was rendered wrong because of the ")" in a species name.
have you considered saturation, or incomplete lineage sorting? The last can also be common in ancient phylogenies if speciation occurred rapidly. Those are natural sources of discordance. You may take incomplete lineage sorting into account by using a species tree method like *Beast and many more. You can also try constraining the trees as a couple of 'true tree' (one for any two main hypotheses you have about species relationships) and then test whether your data fit any of them better than the other.
- Badges
- Science method
- Similar topics
- Biological Science
- Evolutionary Biology
- Phylogenetics
- Phylogenetic Analysis