Bioinformatics Asked by Ace8910 on August 21, 2021
I have been creating phylogenetic trees using RAxML (using Boostrap + ML) and I end up with a bestTree file, outputted in newick format. When I open up the tree in FigTree, I notice that it collapses many of the internal nodes with small values (1 x 10-6). I am working on visualizing the trees using ETE, which when I render the tree includes these small breakpoints. I was wondering if there is a standard for when to collapse nodes that are considered phylogenetically uninformative that I can apply. Thank you for any help.
If 1E6 is genetic distance, this means one nucleotide difference per million base pairs, which is of course incredibly small and bootstrapping at this level doesn't really work. So what I suspect is happening is that the bootstrap is so low it is not present on the tree. You can assign a cut-off bootstrap but anything below 75% is not interpretable in any case. What might be happening is you are producing the bootstrap "cladogram" which is showing a default branch length, but when <1E6 genetic distance it is showing a trifurcation. This simply means the bootstrap of the consensus tree is so low it is collapsing the branch and this is a function of the very small genetic distance at this node.
Looking at your treefile I don't see any collapsed clades.
(Group_9:0.00000100000050002909,(Group_50:0.00507659947127924935, (.
The clade above has not been colapsed it is just a very small distance in a paraphyletic group. It looks fine to me, i.e. Group_9 is 1E6. It could look 'collapsed' simply because it is so small it is difficult to see. I hope thats ok?
Correct answer by M__ on August 21, 2021
Get help from others!
Recent Answers
Recent Questions
© 2024 TransWikia.com. All rights reserved. Sites we Love: PCI Database, UKBizDB, Menu Kuliner, Sharing RPP