Figure 5: Comparative analysis of protein diversity.

(a) Comparison of total CDS length, total Pfam-A domain length and total Pfam-A domain type numbers from the sequenced genomes of a variety of species. All known spliced isoforms were included. (b) Comparison of domain sequence diversity between lancelets and vertebrates. The diversity was directly measured using the numbers of sequence clusters created using BLASTCLUST. All (Pfam-A) domain types and ancient domain types (that is, non-vertebrate-specific domain types) were analysed separately. (b) The increasing trend of average sequence identity of proteins in five sequential phases of the immune response, from recognition to transcription factors. (d) The expansion and diversification pattern of the immune and stress protein gene repertoire. Average protein identity and the number of 1:1 orthologue proteins versus species-specific proteins are shown. (e) The number of novel domain pairs gained by different lineages. Branch length is proportional to the number of novel domain pairs. Numbers outside and within parentheses represent all novel domain pairs and the novel domain pairs containing no vertebrate-specific domains, respectively. Numbers in circles represent the eight important lineages: B. floridae, B. belcheri, amphioxus ancestor, S. purpuratus, deuterostome ancestor, chordate ancestor, vertebrate ancestor and all six vertebrates. More information is provided in Supplementary Notes 11 and 12.