You, me, your pet dog, and any other animal with a backbone are deuterostomes. So are sea stars, sea urchins, and sea cucumbers. During early development you, me, and echinoderms (sea stars et al) are a round ball of cells. The ball of cells invaginates and that opening becomes our anus. This differs from other animals like flies, worms, and snails whose first opening becomes a mouth. A second opening forms later in development, and it becomes our mouth (hence the name “deuterostome”, or mouth second). The deuterostomes can be broken into two groups: the chordates and the echinoderms. The chordates include vertebrates and some less appreciated animals like tunicates and lancelets.

Why am I telling you all of this? Well, a recent paper examines the evolutionary relationship of the deuterostomes, and comes to a surprising conclusion — the tunicates and vertebrates are close relatives, as are the lancelets and echinoderms. In case you didn’t click the links and look at the pictures, the three groups of chordates seem pretty different from echinoderms. Is this a case of conclusions being made using subjective morphological data being overturned by objective molecular data? Of course I’m going to advocate the awesomeness of molecular data, right? Wrong. Molecular data is only better than phenotypic data when the molecular data is good. As Martin at the appropriately named Lancelet has pointed out, there are some long external branches on that tree. Long branches separated by short internal branches are a hallmark of unreliability in phylogenetics. I just want to point out that I’m not a dogmatic molecular phylogenetics supporter — if you don’t understand what you are doing with the data, you can come to some erroneous premature conclusions. I’m not saying that the authors of this study are wrong, simply that I would take their findings with a grain of salt.

This goes in the Nature loves publishing evolutionary genetics papers that seem a bit off file.


  1. #1 Sandra Porter
    August 23, 2006

    Molecular data are great, but only if the sequences alignments are good. And good alignments are hard to get unless the data quality are good (this is why I care about Phred).

    Check out this paper by Xia, X., Xie, Z., and KM Kjer on 18s ribosomal RNA and tetrapod phylogeny. Syst Biol. 2003 Jun;52(3):283-95

    Apparently, people dissed molecular data for 18S phylogenies for years because they were a bit funny, and then these guys came along and found that the phylogenies were based on crappy alignments.