Fig. 5

Identification of the structural conservation of a gene and prediction of CDSs. The method takes known transcripts as input, and then outputs gene and transcript structure models, predicted CDSs and structural orthology relations between CDSs. Top left: the spliced structures of five transcripts known in three species. (1) The known transcripts of a gene allow us to build its intron/exon structure. (2) Pairwise alignments of each functional site and coding block of a gene against another gene sequence allows us i) to identify the orthology between known sites and blocks between two species (orthologous elements are represented in a same column with red lines) and ii) to reveal previously unknown functional sites and coding blocks (dashed box). (3) In each gene, the orthologous coding blocks receive a common name. Renamed blocks of gene j are shown. (4) Revealed elements serve as a basis to predict orthologous CDSs. Here, gene j gained the prediction of the CDS j₂, which is a spliced ortholog of CDS i₂ known in gene i. Indeed, as the ‘C’ coding block and flanking sites (‘ >C]’) were revealed in gene j, the i₂ transcript appears feasible in gene j. No CDS ortholog of the i₂ transcript is predicted in gene k, since the ‘C’ coding element was not found in gene k, making the i₂ transcript unfeasible in gene k