There is a self-extracting archive at this website: http://xungulab.com/software/diverge2/diverge2.html

We want to analyse the GAPDH family.

Dowload the following files (alignment and 3D structure), unzip them and load them into Diverge2:

• gapdh_alignment.fasta.zip
• 2pkq.pdb.zip

In the “Clustering” tab, select “NJ Tree-Making”, “Kimura” method (but you can also produce a tree with ML methods and load it).

• Click on the node at the basis of group C, and reroot the tree.
• Click on the node at the basis of group C, and “Add cluster” ⇒ Name “GapC”.
• Click on the node at the basis of group A, and “Add cluster” ⇒ Name “GapA”.

In the “Gu99” tab, click on calculate. Could take time. It produce the score for the original Type I of functional divergence. In the “Gu2001” tab, click on calculate. Could take time. It produce the score for an updated Type I of functional divergence. In the “Type II Divergene” tab, click on calculate. Could take time. It produce the score for the Type II of functional divergence.

When you click on a column, you can visualise the posterior probability per site.

The visualisation on the multiple seems to cause the program to crash. So use Jalview instead.