Motus Mentis

In a recent post, I recommended that readers with an interest in biology pay a visit to the Tree of Life Web Project, an interactive display of the taxonomic relationships linking all life on Earth. Upon seeing the post, one of our commenters, microbiologist Andrew Staroscik, mentioned that he had rather a low opinion of the ToL website, and, when I asked him if he would explain in more detail, offered a detailed and piquant critique of the way the tree is organized and presented. We are fortunate to have experts such as Andrew among our readership, and rather than let his response languish in the comment thread, where many of you might not have seen it, I am republishing it here.

Malcolm,

… To raise all the things that pop into my head in response to your question would take about 50 blog entries. Saying what I do not like about that tree is easier than offering specifics about what I would prefer to see, but I’ll give it a shot.

Drawing trees like this is difficult but there are two things the ideal tree should show: 1) the true evolutionary distance between different groups of organisms, and 2) the total diversity within each group. Doing this for all of life on one tree is impossible unless you are making a large mural. So there is always a trade-off between 1 and 2. On the main tree introducing a site like this, I’d like to see the actual distances between groups (1) shown at the cost of the total diversity within subgroups (2). Two reasons: first, this is the aspect of the tree least appreciated by most people; second, the main tree is the only place to show overall relationships, while the within group diversity can be preserved in the subtrees elsewhere on the site. The ToL people may disagree, but it is not clear whether they do or not. The tree they chose does not seek compromise, it simply misrepresents both aspects for no good reason.

Here is an interesting anecdote: If you were to have a ham sandwich and a salad for lunch tomorrow, you would be more closely related to everything you are eating (including the yeast in the bread) than all bacteria are to all archaea. The authors of the webpage downplay this point in their choice of that picture. My position is that a webpage dedicated to phylogenetic relationships and the tree of life should be celebrating points like this, not obscuring them. Their calling attention to it would in no way detract from all of the other content on the site.

There is a page on the ToL website explaining their chosen tree. I can use their words to expand on my comments. This is the first paragraph:

“We’ve been getting a lot of inquiries about the tree of life picture on our home page. It is important to note that the major function of this picture is to help visitors to the ToL web site to quickly navigate to pages of some of the major groups of organisms. In order to serve this purpose, we had to use a greatly simplified representation of the tree of life.’

First of all, I take from the tone here that I am not alone in my objections to their picture. They are not really simplifying as much as distorting the tree. I’ve already said that the navigation point is a cop-out. Towards the bottom of the overview page they link to a few sites with other representations. Two I’d highlight as being better are the first (UCMP’s tree) and sixth (Static diagrams from ‘Assembling the Tree of Life’) on Rebecca Shapley’s webpage. The UCMP’s tree in particular shows that there is no problem with using such a tree as a navigation tool. While it is not too pretty, the phylogenetic tree on the wikipedia Monera page is also quite good. Again, I’d argue that groups like ”˜animals’ or ”˜plant’ are neither difficult to find nor hard to click on in these trees.

“Life on Earth shares a common, genetic history with complex origins”¦’

Life does have complex origins and there is a large and rich literature on this topic. Why on earth do they cite an article in the New York Times to make this point? They also say in this section that, due to the complexity, the origins have been left obscure. Maybe I just have a much different idea of what the tree of life is, but why not choose to use their site as a way to explore this issue rather than obscure it.

They also mention horizontal gene transfer (HGT) here. Not only has HGT been important historically, but it continues to be a major factor in the evolution of both bacteria and archaea. No discussion of the relationships among these two domains is complete without a considering HGT. Of particular interest to humans is the role of HGT in bacterial pathogenesis and antibiotic resistance.

The next two sections on this page begin (in red) with “Don’t draw conclusions about the relative diversity of different groups of organisms.’ And “Don’t interpret relative branch lengths as indicators of levels of evolutionary ”˜advancement’ .’

These are essentially my 1 and 2 above and I take these two paragraphs to be acknowledgements on the part of the people behind the site that their tree does neither well.

One sentence I really I object to is this: “Therefore, we created a diagram that is more of a general representation of people’s interest in different groups as well as their coverage on the ToL project.’ Did they just say that my entire research area is not interesting? Is there any wonder why there is little coverage of microbes on the site? What microbiologist is going to take the time to develop content that is going to be ignored?

They repeat essentially the same thing again in the next paragraph saying that they made a conscious choice to relegate a bunch of boring stuff to the bottom corner of the page in order to make room for the really interesting stuff. Jeez, the least they could do is put pictures of microbes over by the other two domains. As it is you have to go through the root to get to bacteria or archaea.

The last red bullet is that “No organism alive today represents the ancestor of any other living creature.’ Then goes on to say: “For practical reasons, we have placed an extant (living) annelid worm smack in the middle of our tree. Of course, this worm is not meant to represent the ancestor of organisms shown higher up in the tree; rather, it stands as a representative of animals in general. The most recent common ancestor of arthropods and vertebrates was some kind of worm, but it certainly wasn’t an annelid; in fact, nobody knows what this creature may have looked like.’

Does this explanation work for a layperson? It does not resonate with me at all. When I look at the picture I don’t see a representative animal, I see a worm in the middle of the tree. No other tree makers that I know of feel the need to stick living organisms on internal nodes and I am not sure why they did. One way other trees succeed in illustrating sub-groups is by color coding branches.

I hope this helps.

Andrew