The two GHMMs representing transcripts. (A) Full transcript model, gray figures represent final states, the arrows represent the flow of the architecture, indicating only one initial state, 5|$^\prime $|UTR. (B) Partial transcript model, gray figures represent final states, the arrows represent the flow of the architecture. The four states represented by circles are states with explicit duration distribution that emit the protein-coding region: fullCDS, pCDS0, pCDS1 and pCDS2. The state fullCDS models a complete coding region. The states pCDS0, pCDS1 and pCDS2 represent partial coding regions that start, respectively, at frame 0, 1 and 2. The state labeled UTR can be used to represent either the 5|$^\prime $|UTR or the 3|$^\prime $|UTR. The 3|$^\prime $|UTR state represents 3|$^\prime $|UTRs. The states Start and Stop in diamonds have a fixed length duration, and they represent the start codon and stop codon, respectively.