Fig. 11. Comparisons of OBP sequences. (A) Amino acid sequence comparisons of 23 D. melanogaster OBPs. A Neighbor Joining Distance tree is shown, derived from the alignment matrix shown in Fig. 9C. Branch lengths are proportional to percentage sequence difference (scale bar represents 10 % mean difference). Three methods were used for this analysis; numbers by nodes are triplets and refer in order to neighbor-joining bootstrap values (5000 replicates), maximum likelihood quartet puzzling support values (in parentheses, 50,000 puzzling steps), and maximum parsimony support values (5000 replicates). Numbers with asterisks indicate gene locus numbers identified in Table 1 and Fig. 9A. Branches are collapsed to 40 % support for at least one method of analysis; all three methods yielded identical topologies at this level of support. Non-drosopohilid taxa are indicated (<>); these were identified when searching the D. melanogaster homologues using Blast. (B) Comparisons of exon domains of 23 D. melanogaster OBPs. Graphical representation of aligned amino acid sequences shown in Fig. 9C, focusing on the alignment of exon domain boundaries within the proteins, including 23 D. melanogaster OBPs plus PBP1Msex and GOBP2Msex (transferred from Fig. 4). Alternate exon domains are shown as filled and unfilled boxes; C-terminal amino acid numbers of exon domain boundaries are indicated, referencing their character positions in the alignment (Fig. 9C). (C) Amino acid sequence comparisons of dipteran versus M. sexta OBPs. A Neighbor Joining Distance tree is shown (Paup 4.0b8), derived from a ClustalX alignment (not shown). Branch lengths are proportional to percentage sequence difference (scale bar represents 10 % mean difference); numbers by nodes are bootstrap values (1000 replicates). Sources of dipteran sequences are described above, except for the mosquito sequences; two sequences from Anopheles gambiae are from L. Zwiebel, and one sequence from Aedes aegypti is from J. Bohbot and R. Vogt. M. sexta sequences PBP1, PBP2, PBP3, GOBP1, GOBP2, ABPX and ABP1 were previously published (Györgyi et al., 1988; Vogt et al., 1991b; Robertson et al., 1999). The remaining M. sexta sequences were identified from ESTs submitted by H. Robertson (GenBank); this data set was downloaded and searched locally by Blast protocols using software obtained from the NCBI FTP site, and subsequently translated for alignment. Representative EST accession numbers are indicated. One sequence, ABP4, was provided by Hugh Robertson and is as yet unpublished. The broken bar at the bottom identifies major similarity groups in this analysis.