Figure 3. Deriving Orthology via Transcript Phylogeny.

(A,B) Phylogenetic relationships for a dog (D₁) and three human (H₁, H₂, and H₃) genes. D₁ is the orthologue to H₁ and H₂. H₃ has been orphaned by the loss of its dog orthologue. Each gene has two splice variants A and B (B), and their transcripts are subscripted accordingly.

(C) Phylogenetic relationships for all transcripts. Each group or clade of orthologous transcripts recapitulates the gene orthology in (A). The transcripts A and B for the orphaned gene H₃ are also themselves orphaned on the transcript tree. The transcripts from clade 1 are selected to represent the three genes (D₁, H₁, and H₂) because phylogenetic distance between orthologues (arrow 1) is smaller than that for clade 2 (arrow 2).

(D) How orthology is predicted when transcripts are missing. D_1A and H_2A are selected as the representative transcripts for their genes because the d_S between these orthologues is smaller than that for D_1B and H_1B. The transcripts in clade 1 are used to predict orthology between D₁ and H₂. Though H₁ also has transcripts in orthologous relationships with D₁, orthology between these two genes is not predicted, leaving H₁ as an orphan. No orthology predictions are made for the gene H₃, which remains as an orphan.