Mj. Donoghue et S. Mathews, DUPLICATE GENES AND THE ROOT OF ANGIOSPERMS, WITH AN EXAMPLE USING PHYTOCHROME SEQUENCES, Molecular phylogenetics and evolution (Print), 9(3), 1998, pp. 489-500
The root of the angiosperm tree has not yet been established. Major mo
rphological and molecular differences between angiosperms and other se
ed plants have introduced ambiguities and possibly spurious results, B
ecause it is unlikely that extant species more closely related to angi
osperms will be discovered, and because relevant fossils will almost c
ertainly not yield molecular data, the use of duplicate genes for root
ing purposes may provide the best hope of a solution. Simultaneous ana
lysis of the genes resulting from a gene duplication event along the b
ranch subtending angiosperms would yield an unrooted network, wherein
two congruent gene trees should be connected by a single branch. In th
ese circumstances the best rooted species tree is the one that corresp
onds to the two gene trees when the network is rooted along the connec
ting branch. In general, this approach can be viewed as choosing among
rooted species trees by minimizing hypothesized events such as gene d
uplication, gene loss, Lineage sorting, and lateral transfer. Of those
gene families that are potentially relevant to the angiosperm problem
, phytochrome genes warrant special attention. Phylogenetic analysis o
f a sample of complete phytochrome (PHY) sequences implies that an ini
tial duplication event preceded (or occurred early within) the radiati
on of seed plants and that each of the two resulting copies duplicated
again. In one of these cases, leading to the PHYA and PHYC lineages,
duplication appears to have occurred before the diversification of ang
iosperms. Duplicate gene trees are congruent in these broad analyses,
but the sample of sequences is too limited to provide much insight int
o the rooting question. Preliminary analyses of partial PHYA and PHYC
sequences from several presumably basal angiosperm lineages are promis
ing, but more data are needed to critically evaluate the power of thes
e genes to resolve the angiosperm radiation. (C) 1998 Academic Press.