Information
This command-line Java program allows users to ‘pare’ down their tree by either removing unwanted leaves (tip-nodes), removing bootstrap information from the tree, or removing branch lengths from the tree – or any combination. Both of these functions can be accomplished in languages like R or Perl, but Java allows very large trees to be pared down quickly, efficiently, and easily!
The program takes Nexus/Newick-style phylogenetic tree code as input that may or may not include bootstraps, and outputs Newick-style trees with the requested nodes and/or bootstrap values removed.
You can read more about the program below ('Program Details') and in the manual ('Program Links').
Remember: Deleting nodes from phylogenies is risky business! See 'Program Details' for more.
Bug? Question? Comments? Feedback form
Program Links
PLEASE NOTE: I cannot be held responsible for any incorrect behaviour of this program.
Though I try to thoroughly test all my programs, I cannot test them in every conceivable
situation, and so cannot guarantee that it will behave correctly or as predicted at
every execution.
Download Jar File
User Manual
Program Details
It is worth mentioning that the total root-to-tip distance is kept the same for every leaf after other leaves are deleted. (Non-deleted nodes will have the same total distance to the root as they did before. See the end of the manual for more details.)
Updates Since v1.0.1: (List of previous updates at the bottom of the page)- ‘-topo’ can now be used to specify that output files should be written without branch lengths (topology only). Can be combined with '-nbs' and '-del'/'-keep' or used alone.
- Now compatible with Windows Shell auto-glob when passing * to specify file endings
- ‘-t’ is now optional - if not included the value defaults to 'O'
- ‘-rax’ is now depreciated (but can be included without affecting the run). The program now automatically detects and handles a larger number of minor variations in format that can occur at the end of Newick files
Reminder:
Deleting nodes from phylogenies is risky business! Deleting nodes is unlikely to give you
the same phylogeny as re-running the tree with the corresponding sequences removed. You could
easily misrepresent your data to yourself and to others by deleting nodes to make a ‘prettier’
picture. Thus, I implore the user to run this program carefully and I do not condone it being
used to intentionally misrepresent data!
Here are a few (safer) situations where this program might be useful:
-
Deleting identical sequences
Some phylogenetic programs do not warn of identical sequences at the beginning of a run, leading to leaf-pairs with branch-lengths of 0. Some programs have trouble with or will not accept branch lengths of 0. If the tree is large and the run takes many hours or days, deleting the identical sequences and re-running the tree (or trees) can be a very unappealing task. Because both leaves have a branch length of 0 from the previous node, deleting one will not affect the legitimacy of your tree, and can save a lot of time. Also, the program can be put into a batch file or script so that the same sequences are deleted from many tree files quickly and easily. -
‘Zooming In’ on Trees / Simplifying Outgroups
For display purposes, PareTree could be used to delete all but a cluster of sequences from a tree, allowing a better, ‘zoomed-in’ image of a section of the tree the tree for easier examination. Similarly, if multiple, monophyletic outgroup sequences have been used, but are affecting the display of the tree, all but one could be deleted (and then this one renamed to reflect that it was originally a group of sequences).
Citation
If you publish or present work that has been processed using this program, please
cite Emma Hodcroft and the website where this program can be downloaded
(http://emmahodcroft.com/PareTree.html).
(I'd also be very interested to hear about what you've used the program for, so feel free to send me a link to your paper/research!)
Bug? Question? Comments? Feedback form
Previous Updates
Updates Since v1.0:- Corrected file/directory reading for Unix/Linux/Mac users
- Corrected error that assumed all files had a path before the file name
“To iterate is human, to recurse divine.”
