GOAD.DAVISON@BIONET-20.BIO.NET (Dan Davison) (02/16/89)
Several recent articles are worth noting for their coverage of just how far one can take evolutionary data . David Penny has a News and Views article in a recent Nature (26 January 1989; vol. 337:304-5) titled "What, if anything, is _Prochlorion_?". This article is an overview of two articles in the issue (p. 380 and 382) which come to conflicting conclusions as to the phylogenetic position of _Prochlorion_. One group (Turner et al., p380) compared the small subunit ribosomal RNA sequence with other small subunit rRNA sequences, and the other (Morden and Golden, p.382) a part of the photosystem II protein complex. Penny says: "The biggest difficulty is in persuading people that collectin sequences is the easy part of any study. The hard part is estimating the accuracy of the results...Any measurement, to be scientific, must include an indication of its accuracy. Despite the progress that has been made, reconstruction of evolutionary trees from sequence data must still appear to those in other fields as parascientific." On the same subject, there are a series of letters in the "technical comments" section of Science, 27 January 1989, pp. 548-551. The letters were inspired by "Molecular phylogeny of the animal kingdom", Field et al., Science 239:748 (1988). The trees presented in that article have a number of peculiarities, and these are discussed by the correspondents. There is also a reply by Field et al. These three articles once again puzzle me. Why do investigators persist in examining one trait and claim a conclusive phylogeny? The case for using rRNA sequences is again reiterated by the Field et al. response (absolute requirement for life, high resistance to confusion due to the small likelihood of successful lateral gene transfer, among others). The problem still remains that (IMHO) there is no subsititute for multivariate analysis. (Such multivariate analysis was discussed in this list by Kramer, Claverie, and Felsenstein in the late spring of 1988). dan davison theoretical biology and biophysics t-10 ms k710 los alamos national laboratory los alamos, nm 87545 USA dd@lanl.gov (internet) dd%lanl.gov@CUNYVM (bitnet) ...cmcl2!lanl!dd (uucp, maybe) -------