Generate a phylogenetic tree

MegAlign Pro offers four tree-building algorithms:

• Neighbor joining: BIONJ uses the BIONJ algorithm (Gascuel, 1997), a variant of the Neighbor-Joining algorithm (Saito and Nei, 1987) that was used in classic MegAlign. Choose this option if your project includes fewer than 4 sequences, the minimum needed to use the RAxML methods below. Otherwise, a RAxML method is recommended: particularly if your data set is divergent, contains many taxa and/or very long sequences, or if you need to enter support (bootstrapping) values.
  
  
• Maximum Likelihood: IQ-TREE was developed by Nguyen et al. (2015). See the IQ-TREE website for in-depth user documentation and article citations. Some published reviews note that iterations of IQ-Tree may be faster and/or more accurate than RAxML. We recommend trying both and judging based on your specific data set. DNASTAR ran some informal speed tests using a data set consisting of 250 sequences and 1,000 replicates. Without bootstrapping, IQ-Tree generated a tree in 15 minutes. With bootstrapping set to 1,000 max. iterations, IQ-Tree completed quickly, taking only 26 minutes for the same data set.
  
  
• Maximum likelihood: RAxML was developed by Stamatakis (2014). Comprehensive information about options associated with this algorithm can be found in the RAxML Manual (PDF).
  
  
• Maximum likelihood: RAxML-NG was developed by Kozlov et al. (2019) and was added to MegAlign Pro version 17.6, released in spring 2024. This next generation algorithm is faster and has a larger capacity than its predecessor, RAxML. This algorithm can be used for any data set with enough sequences, but is particularly useful for gene homology alignments, as it supports the creation of very large trees.

For helpful advice on options related to phylogenetic tree building, see this December 2020 DNASTAR blog post by Brian Walsh, PhD. This post pre-dates the addition of RAxML-NG and IQ-TREE to MegAlign Pro, but still contains valuable tips.

To generate a new phylogenetic tree:

1. Add sequences and perform a multiple alignment.

2. Use any of the following methods to choose your preferred tree-building algorithm.
   
   
   • Click on the Tree tab to open the Tree view, then click on the desired method.
     
     
     
     
   • In the Trees section of the Explorer panel, press the green plus-sign tool () to the right of the table. In the Using drop-down menu, select the desired algorithm.
     
     
   • Use the menu commands Tree > Compute Phylogeny Using > Neighbor Joining:BIONJ, Tree > Compute Phylogeny Using > Maximum Likelihood: IQ-TREE, Tree > Compute Phylogeny Using > Maximum Likelihood: RAxML, or Tree > Compute Phylogeny Using > Maximum Likelihood: RAxML-NG.

3. Customize settings based on the algorithm you chose.
   
   
   • If you chose Neighbor joining: BIONJ:
     
     
     
     
     • Type in a new Name for the tree or keep the default name of Tree <number>.
       
       
     • If you change your mind and want to use the other algorithm, choose it from the Using drop-down menu.
       
       
     • Press OK to generate the tree.
       
       
   • If you chose any Maximum Likelihood method:
     
     
     
     
     • Type in a new Name for the tree or keep the default name of Tree <number>.
       
       
     • If you change your mind and want to use the other algorithm, choose it from the Using drop-down menu.
       
       
     • To add a bootstrapping step to the calculation, place a checkmark next to Bootstrap analysis; if you do so, another option appears. Type in the number of Iterations or keep the default value of 100. Note that a bootstrap value of 50% or less is equivalent to a coin flip. For trustworthy results, we recommend using a 70% cutoff or higher.
       
       
     • Keep the default Seed value, type in a new integer, or select Choose > Random to enter a random value in the box.
       
       
     • Enter the desired number of Threads. The default value is 2.
       
       
     • Press OK to generate the tree.

4. (optional) Use the Style panel’s Tree section to customize the tree as desired.

5. (optional) Create another phylogenetic tree using a different algorithm or different settings for the same algorithm. See Work with multiple trees.