Figure 2

Annotation of a simulated alignment by using PRANK 's genomic structure model. A. A schematic representation of the structure of the genomic region used in the simulation. Noncoding regions, evolving by a K80 (Kimura two parameters) substitution process [32], are shown in green. Coding regions are shown in blue, and evolve by a GY94 (Goldman-Yang) codon model [33, 34]. The other features included in the simulation, the fixed start codon and splicing sites and the stop codon evolving by a special substitution process, are shown in lighter shades. B. A "true" multiple sequence alignment resulting from the simulation of the genomic region along the phylogenetic tree shown to the left. The tracks under the sites represent the true intron-exon structure ("True state") and the annotation of the alignment inferred by PRANK alignment tool transferred to the human sequence ("Prank annot."). The thin portions of the PRANK annotation track indicate positions that have no annotation available as they have gaps in the human sequence in the true simulated alignment.