i have made a phylogenetic tree to show the relationships betwen the ITS sequences from different Enterocytozoon bieneusi genotypes. I also have inputed a sequence obtained by our laboratory thats identical to just one sequence of those genotypes. But at the phylogenetical tree they appear as different leaves separated by a significant percentage of bootsptraps. Obviously im pretty new at this.
genotype D and BsAs1 are identical sequences! at .jpg tree
You need to describe how you inferred the tree so that the problem may be diagnosed.
What kind of tree building method did you use? Identical sequences must pair up.
thanks Patrice Showers Corneli Patrice to answer, before doing the phylogenetical analisis i aligned the sequences ( 100 different ITS sequences moreless from 100 different Genotypes from GeneBank) and my ITS sequence (BsAs1). all of them where te same 243 nucleotide lenght ITS sequences, and i aligned them by Muscle in Mega software. Then i ran a model test to find the best model, and i utilized Maximum likelihood as method with HKY-G model, with 1000 bootstrap.
my sequence called BsAs1 has 100% identity with "Genotype D" , so both sequences are identical, so my sequence is a genotype D, but my issue is why when i align and build a phylogenetical tree they appear as different "leaves" ?
Very nicely done analysis. When you infer a tree with the implicit assumption of bipartitioning, you will not get two taxa - even if identical - on the same branch. The corresponding bootstrap values are in no way contra vincing. At the root of the clade is a very unconvincing 36% bootstrap and the subsequent values in the 80s are in no way strong evidence for the separation into two clades. None of the relationships among taxa in that part of the tree are well supported..
I suggest that you post the ML tree with proportional branch lengths rather than as a cladogram which exaggerates the evolutionary distance among the sequences. I am guessing that doing so you will find that extremely short branches join the taxa in this clade. Very short branches indicate how very similar the sequences are.
These data may be insufficiently phylogenetic informative to resolve that part of the tree. That is those taxa may be too similar to reveal differences.
Should you wish to include your alignment, I can take a look and diagnose the reason that D and BsAs1 are not appearing as sister taxa. I'd be happy to do so as this is a very good question from you and reveals a common misconception about the information the realtionship between the bootstrap and the all-important branch length data.
Thanks you very much Patrice!!! so clear!! here i left the alignment as fasta file. Im resolving to drop my BsAs1 Secuence from the analisis and describe that it is an allready known genotype called D, and that its position at the philogenetical tree is the same position of this D genotype. Just to avoid that misconception you talked about. what do you think?
Hi Cristian,
I made some trees using the HKY+G model in PhyML (part of the free Seaview package) using your fasta alignment. Both are presented with the branches proportional to the ML estatimate. The first uses all taxa but the branch to the two apparent outgroups is so long that it makes the rest of the tree less readable. These two are clearly very different from all the others. The second tree leaves them out for clarity.
The tree that you included is called a cladogram because the branch lengths are arbitrary. The idea is just to get a notion of the branching pattern. You can see that a cladogram can be very misleading.
These trees show the branching pattern and the branch lengths which are proportional to the number of substitutions on each branch. This so-called phylogram gives a clear picture of the rate of evolution on each part of the tree.
Let me know if you have any questions at all.
Oops! I sent the same tree twice. Here is the more readable one with support values that are at least reasonably large. Many zero branch lengths identical or nearly identical sequences. Keep them all anyway if the samples are are from different specimens. Zero branch lengths are still informative.
Thanks you so much Patrice!!! i struggled looking for this info at books and i didnt find any answer as yours!
- Badges
- Science topic
- Similar topics
- Linguistics
- Composition Studies
- Publications