02 February 2009

Names on Nodes Revisions

If you look at the taxonomy in this earlier post, you might notice something odd (although you'd have to look pretty hard and be familiar with certain literature). Namely, Palaeognathae and Panpalaeognathae are shown as equivalents. While they are the same in terms of composition in this context (i.e., their "finest" members are Struthio camelus, Tinamus major, and their final common ancestor), the former is a crown group and the latter is a total group, so they are unlikely to be actual synonyms.

To get around this, I changed the Newick tree interpretation algorithm to insert an extra ancestor, called a "branch ancestor" in the middle of every arc. A branch-based definition (such as for a total group) will include this, but a node-based definition (such as for a crown group) will not. Voila, they can be distinguished. (This still doesn't help to distinguish apomorphy-based definitions, but, honestly, I'm not about to bend over backwards for apomorphy-based definitions.)

I added some more definitions, but it was proving to be too taxing on memory (the computer's, not mine). I had to make a major optimization in the way that relations are determined to fall under a certain context. Seems to be working better so far.

Using two trees at once was unfortunately causing problems, so I took the species-based one out. I'm pretty sure it would work, though, if I had better ways of equating hypothetical ancestors between different trees. That's something I'll be focusing on next.

The latest version of the generated taxonomy:

No comments:

Post a Comment