More on Schwartz and Orangutans

I haven’t had a lot of time for reading lately, so I am still in the process of reading the paper. I am just now at the critique of the molecular evidence for a chimp/human clade. Here is a longish quote:

Morphological data are not immune to the possibility of homoplasy, but neither are molecular data. In fact, the lack of a comprehensive theory of homology in molecular systematics is its main weakness. As with demonstrated morphological similarity, similarity between DNA sequences can be due to primitive retention, reversal, homoplasy or simply to being non-homologous (e.g. convergent). The only bases for claiming demonstration of molecular homology are the limited data set of four nucleotides and their positions relative to each other (sequence order). DNA sequences are further inherently ambiguous because a substitution event leaves no evidence of replacement, which would seem to be a critical element towards hypothesizing whether matching base pairs represent primitive retention, convergence, or unique derivation.

In addition, in order to compare supposedly homologous DNA sequences one must align sequences of different lengths, which is a procedure that requires assumptions about deletions or additions that underlie the observed disparity in nucleotide sequence order and length. In the end, there is no objective way to assess the relative phylogenetic value for the number of gaps and substitutions that are assumed in order to align sequences of different lengths (Marks, 2003). Thus, statements of sequence homology are not generated from individual comparative outgroup character analysis as they are in morphological analyses. Rather, the claim of sequence homology is the result of an overall best fit between an artificially reconstructed sequence and subsequent measures of phenetic
similarity (Giribet et al., 2002; De Laet, 2005; Redelings & Suchard, 2005; Phillips, 2006; Kjer et al., 2007).

Discuss amongst yourselves…

8 Responses

  1. Was does the phrase “special pleading” keep coming to mind?

  2. From National Geographic:

    “There are many paleontologists and molecular biologists who are heaping scorn on this paper,” noted Peter Andrews of the Natural History Museum in London.

    It is not a very strong paper. Although it mentions 28 traits that humans share with Orangs, they are not discussed in the paper. You have to look in the supplementary material to even find out what they are. The section on molecular biology wasn’t that great and I haven’t made up my mind about the biogeography section.

  3. Have you asked ERV to comment? That first paragraph is just pleading for some of her special brand of scorn.

  4. No, actually, the thought had not occurred to me until you mentioned it.

  5. Just reading this exceprt I’m not impresed. The author’s knowledge of modern molecular phylogenetics is poor. While some of the issues such as the impact of alighment quality on phylogenetic reconstruction are something we are concerned about (and acknowledge) his entire critique apears limited to nucleotide alignment and analysis. While nucleotide level analysis is important for resolving trickier relationships amino acid based alignments are preferable (if it is appropriate to do so of course) not to mention codon models of evolution. While some molecular phylogenetic reconstructions leave much to be desired (an alignment in just CLUSTAL and a NJ tree? Really?) there are a great many experts, used to working with much older divergences, who have contributed greatly to molecular systematics. Highly sophisticated statistical tools do exist and are routinely used. Alignment editing may sometimes be more of an art than a science we do, again, have excellent statistical tools to guide us. We can also be guided by knowledge of protein domain, secondary structures, etc.

    I am extremely underwhelmed by the critique of molecular systematics.

  6. That is the highpoint of the paper.

  7. Just an question here. I’m not in this field, so forgive me if I am way off base here.

    I just looked at this article: (from a link on afarensis)

    To my eyes, at least, Anoiapithecus brevirostris looks more like an orangutan than a chimp or gorilla. It’s from 12 million years ago. Heh-maybe it’s the hair?

    So let’s say the paper’s molecular conclusions are incorrect, but the morphological analysis is pretty good.

    Does current evidence preclude the possibility that the last common ancestor of african apes and orangutans (about 14 million years ago) may have been very “orangutan-like”, or even that the last common ancestor of the chimp/human/gorilla clade may have been also?

  8. Good question. Currently, one would expect the common ancestor of the gorilla/chimp/human clade to look something like Pierolapithecus and the common ancestor of that clade with orangs to look something like Hispanopithecus or perhaps Ankarapithecus…

Comments are closed.

%d bloggers like this: