Phylogenetic analyses of 110 serpin protein sequences revealed clades consi
stent with independent phylogenetic analyses based on exon-intron structure
and diagnostic amino acid sites. Trees were estimated by maximum likelihoo
d, neighbor joining and partial split decomposition using both the BLOSUM 6
2 and Jones-Taylor-Thornton substitution matrices. Neighbor-joining trees g
ave results closest to those based on independent analyses using genomic an
d chromosomal data. The maximum-likelihood trees derived using the quartet
puzzling algorithm were very conservative, producing many small clades that
separated groups of proteins that other results suggest were related. Inde
pendent analyses based on exon-intron structure suggested that a neighbor-j
oining tree was more accurate than maximum-likelihood trees obtained using
the quartet puzzling algorithm.