Applied Bioinformatics 2003:2(3 Suppl) xxx–xxx © 2003 Open Mind Journals Limited. All rights reserved. 1 ORIGINAL RESEARCH AUTHOR PROOF COPY ONLY Introduction Tetrapods (four-limbed vertebrates, including humans) are generally accepted to have evolved from Sarcopterygii (lobe-finned fish), rather than Actinopeterygii (ray-finned fish), although there is weak support for the latter option (Baba et al 1984; Bihop and Friday 1988). The two extant super-orders of lobe-finned fish are the coelacanth (Coelacanthimorpha) and lungfish (Dipnoi). The phylogeny of lobe-finned fish, including the clustering of tetrapods, is still unresolved, mainly, because the transition from fish to tetrapod seems to have occurred within a very short time- frame of only 10–20 million years, approximately 400 Mya (Maisey 1986; Carroll 1988; Cloutier and Ahlberg 1996). Phylogeneticists essentially have the problem of trying to focus on something very small that is far away, often resulting in short branch lengths and inaccurate trees. There are three possible schemes of tetrapod evolution in relation to extant lineages (Figure 1), with tree (B) historically preferred. Today, however, there seems to be little evidence forthcoming to support a close coelacanth/ tetrapod relationship (eg Frtizsch 1987; Gorr et al 1991; Schultze 1991; Stock et al 1991; Waehneldt et al 1991; Gorr and Kleinschmidt 1993; Bogart et al 1994; Tamai et al 1994). Rosen’s radical 1981 paper, proposing that lungfish were the direct ancestor to tetrapods, provided the seed for further lobe-finned fish phylogenetic studies (Rosen et al 1981). Since then, there has been much support for the less radical proposal that lungfish are the closest living relatives to tetrapods (Figure 1A), with evidence from morphological studies (Ahlberg 1991), physiological studies (Hughes 1998), protein analysis (Mommsen and Walsh 1989; Noso et al 1993; Hansen and Hansen 1994; Vallarino et al 1998), nuclear-encoded gene analysis (Hyodo et al 1997; Tohyama et al 2000) and mitochondrial analysis (Meyer and Wilson 1990; Meyer and Dolven 1992; Hedges et al 1993; Zardoya and Meyer 1996b; Cao et al 1998; Rasmussen et al 1998; Rasmussen and Arnason 1999). The third tree (Figure 1C) clusters coelacanths and lungfish as sister taxa that are equally related to tetrapods. There has been a recent burst of support for this tree, which has arisen from anatomical data (Mee-Mann 1991; Schultze 1994) and mitochondrial analysis (Yokobori et al 1994; Zardoya and Meyer 1996a). With all tree possibilities supported, how does one determine which one is correct? Zardoya, Meyer and others have performed most of the phylogenetic studies involving mitochondrial sequences (see Zardoya and Meyer 1996b; Zardoya and Meyer 1997a; Zardoya and Meyer 1997b; Zardoya et al 1998; Zardoya and Meyer 2000). These studies, using complete mitochondrial genome sequences, either on their own, or in conjunction with 28S sequences, are the most thorough analysis of extant lobe-finned fish relationships yet. However, all their recent studies have found that different phylogenetic methods were unable to resolve these relationships, with either a lungfish + tetrapod Correspondence: Victoria Metcalf, Department of Zoology, University of Canterbury, Private Bag 4800, Christchurch, New Zealand; tel +64 3 3 364 2987 ext 7019; fax +64 3 364 2024; email victoria.metcalf@canterbury.ac.nz Using serum albumin to infer vertebrate phylogenies Victoria Metcalf, 1 Stephen Brennan, 2 Peter George 2 1 Department of Zoology, University of Canterbury, Christchurch, New Zealand; 2 Molecular Pathology Laboratory, Department of Pathology, Christchurch School of Medicine, Christchurch, New Zealand Abstract: The phylogenetic relationships of the two extant lobe-finned fish, coelacanth and lungfish, to tetrapods remain unresolved. In our study of serum albumin in lungfish we determined 101 residues of protein sequence from five peptides. The lungfish sequences were aligned with the albumin superfamily using the UPGMA-based pileup or the progressive ALIGN programs. The tree construction methods employed were maximum parsimony, neighbour joining and TREE. The lungfish albumin sequence was compared with that of other species, and the resulting trees indicated a close relationship of lungfish to tetrapods, in line with the bulk of current phylogenetic evidence. Keywords: albumin, lungfish, tetrapod, coelacanth, phylogeny