ConsensusBranchLengths is a module for Mesquite which calculates mean branch lengths for consensus phylogenies. It was primarily developed to determine branch lengths for consensus trees obtained from BayesPhylogenies analyses, but it will work with any other source of trees (e.g. MrBayes output, etc.).
Features of the module:
- Calculate mean branch lengths for consensus phylogenies
- Determine standard deviation and 95% CI of branch lengths
ConsensusBranchLengths does not calculate the consensus tree topology itself. The module is still in early development and should be considered beta software. It is freely provided here, but I will not take responsibility for data loss or other damage that results from its use. ConsensusBranchLengths should work on any platform that Mesquite supports (Mac OS X, Linux and Windows). Please contact me with questions, comments, bug reports, or feature suggestions!
ConsensusBranchLengths v0.1 (8 KB)
- Download and install Mesquite
- Download ConsensusBranchLengths.zip
- Unzip the file and place the resulting "schmidt" folder in Mesquite_Folder/mesquite/
- Launch Mesquite and open your data file
- Create a new tree window with your consensus tree
- With the focus on this tree window, go to the Tree menu and select Alter/Transform Branch Lengths > Consensus Branch Lengths
- You will be asked to select a source of trees (e.g. the set of post-burnin trees sampled by BayesPhylogenies). The trees can be read directly from a file, but for best performance (especially with large files) it is best to import the sampled trees ahead of time (Taxa&Trees > Get Files With Trees > Include Partial Contents..., and specify the burnin).
- Important: ConsensusBranchLengths currently requires that both your consensus tree and your sampled trees be rooted. You will get strange results if this is not the case. I am working on resolving this limitation.
- ConsensusBranchLengths will now go through the tree set, and for each branch in the consensus tree it will calculate the mean length of that branch among all sampled trees that contain the branch. Trees which do not have that branch are ignored. It will also calculate the standard deviation and 95% confidence interval for each branch's length.
- The branch lengths of the consensus tree will be dynamically updated to reflect the newly calculated branch lengths.
- The branch number, number of sampled trees with the branch, mean branch length, standard deviation of branch length, and 95% confidence interval for each branch will be printed to the log.
- The branch number corresponds to the number of the node at the apical (furthest from the root) end of the branch. You can determine the node number by selecting Drawing > Show Node Numbers.
- If a branch in the consensus tree does not occur in any of the sampled trees (which shouldn't happen in normal situations, but could happen if you've modified the consensus topology, for instance), that branch will be given a length of zero, and will be indicated as such in the log.