Things I didn’t know about phylogenetic analyses based on DNA molecules.

In my never ending quest to understand reptile phylogeny
I was fortunate to read Scotland et al. (2003) and Jenner (2004). Thankfully the latter rebutted the former. Scotland et al. are all plant scientists, so bear in mind, they deal with far fewer ‘moving parts’ in the taxa they study.

Scott et al. (2004) wrote: “We present the view that rigorous and critical anatomical studies of fewer morphological characters, in the context of molecular phylogenies, is a more fruitful approach to integrating the strengths of morphological data with those of sequence data. This approach is preferable to compiling larger data matrices of increasingly ambiguous and problematic morphological characters.

“In conclusion, problems surrounding character coding of morphological data reduce the number of unambiguous morphological characters for analysis. The crucial issue for morphology is that the already small number of morphological characters is further compromised by ambiguous homology assessment.

“DNA is much simpler. There is no ambiguity that the unit of comparison is the nucleotide and that adenine, guanine, cytosine, and thymine represent different versions of the same entity.
 
“Hillis and Wiens (2000) stated that dense taxon sampling is the greatest advantage of morphological data, citing recent simulation studies demonstrating the importance of taxon sampling for accurate phylogeny estimates (Hillis, 1996, 1998; Graybeal, 1998). For example, in one simulation study, Graybeal (1998) demonstrated that under some conditions phylogenetic accuracy was improved as the number of taxa increased, but not when more characters were added.”
There it is!. That’s what I’ve been saying!
Here’s the main problem with too few characters
according to Scotland et al. 2003):

“Another important issue relative to increased taxon sampling, in the context of morphological data, relates to the potential decreased number of unambiguous charactersas more taxa are added to a study. Characters that were discrete [in smaller studies] are no longer discrete when additional taxa were added.”
What the large reptile tree tells us:
Discrete characters are fine (they were Larry Martin’s favorite subject). But they’re not important in the scheme of things. What is important, as we’ve always heard, is the suite of characters present in each taxon. Let’s face it, sister taxa share all the characters that lump them together, except for the few that split them apart. And that happens again and again at every one of the 415 nodes in the large reptile tree.

A raft of clarity from Jenner 2004.
Jenner argued against Scotland et al. (2003) like this: “Scotland et al. (2003) evaluated the role of morphology in phylogeny reconstruction, and concluded that morphological evidence offers no hope to resolve phylogeny at any taxonomic level. Consequently, they advocated a very restricted role for morphology in phylogenetics, mainly by mapping selected morphological characters onto molecular phylogenies. I critically examined the scientific basis for the arguments of Scotland et al. (2003), and found them to be unconvincing.”

This is most enlightening from Jenner 2004:
“Nucleotides are characters of relatively low complexity, and the character state space for nucleotides is much more restricted than for morphology. In certain circumstances this creates a considerable danger that the same nucleotide has evolved independently in the same position, and this realization has been an incentive to develop models of evolution that estimate the probability that the same nucleotides at a site are historically identical, and to explore the value of more complex molecular characters. In contrast, morphology generally presents a richer space of more complex characters, which allows a more fine-grained comparison of potential homology, and this may help explain why in certain cases morphology may be qualitatively superior to molecules when considered per character.

“Scotland et al. (2014, 541) claim that these problems of “subjectivity and interpretation” are absent from molecular data, because “areas of ambiguity [in sequence alignment] can be excluded.” As recent research shows, to choose this way of least resistance may be thoroughly misleading, and this short statement seriously underplays the degree of subjectivity and interpretation asocial ted with molecular phylogenetics.”

Jenner then discussed more than a decade of 18S rDNA studies that suggested bird/mammal affinities, which, of course, was in conflict with morphological studies and other molecular data. Jenner continued:

“After the 18S data was analysed in various different ways by different workers, they concluded that this was an example of different molecules giving significantly different estimates of phylogeny. However, a recent study by Xia et al. (2003) convincingly showed that the conflict between 18S data and the traditional and other molecular data was an artifact attributable to two main factors: misalignment of sequences, and inappropriate estimation of base frequency parameters.

“Crucial to the resolution of this paradox was the incorporation in the molecular data set of those regions of the 18S molecule that were most variable, and most difficult to align unambiguously. This study clearly showed that restricting the data set to only the least unambiguous sites might produce a thoroughly misleading phylogeny. The problem that ‘different workers will perceive and define characters in different ways’ is therefore certainly not limited to morphological data.”

Ater reading Jenner (2004), you won’t wonder about DNA studies anymore. They’re not perfect and may never be. They don’t work for fossil taxa (you knew that already) and they often come up with bizarre results.

References
Graybeal A 1998. Is it better to add taxa or characters to a difficult phylogenetic problem? Systematic Biology 48:9-17.
Hillis DM 1996. Inferring complex phylogenies. Nature 383:140- 141.
Hillis DM 1998. Taxonomic sampling, phylogenetic accuracy, and investigator bias. Syst. Biol. 47:3-8.
Hillis DM and Wiens JJ 2000. Molecules versus morphology in systematics. Pp 1-19 in Phylogenetic analysis of morphological data (J. J. Wiens, ed.). Smithsonian Institution Press, Washington, D.C.
Jenner RA 2004. Value of morphological phylogenetics. Accepting Partnership by Submission? Morphological Phylogenetics in a Molecular Millennium. Systematic Biology 53333-359.
Scotland RW, Olmstead RG and  Bennett JR 2003. Phylogeny Reconstruction: The Role of Morphology. Systematic Biology 52:539-548.
Xia X, Xie Z and Kjer KM 2003. 18S ribosomal RNA and tetrapod phylogeny. Systematic Biology 52:283-295.

Advertisements

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s