Fig. 2

Genome diagrams of Porphyromonas gingivalis phages highlight functional annotations and gene order conservation in three large clades defined by distinct use of host genome insertion sites. Representations of Pg phage genomes (30 full, 5 partial; names of full-length phages are in saturated colors and partial phages are in lighter shades; “b” suffix indicates version of an “a” phage found in a different assembly of the Pg strain), generated using Clinker [42] and showing predicted protein-coding genes as block arrows colored based on predicted protein functional categories (see Supplementary Fig. 7 for version with protein clustering). Relationships among Pg phages shown in midpoint-rooted tree at left, based on whole genome nucleotide BLAST distance and scaled by VICTOR [28] d0 formula (recommended for nucleic acid datasets). Candidate genus- and species-level clusters are shown for full-length phages in the yellow bars. Three higher-order clades of phages defined by distinct insertion sites in host genomes (by full-length phages only) are highlighted by coloring of phage names (orange: transposition-based insertion; purple: tRNA-ser; green: tRNA-pro). White stars mark phage genome ends defined by contig ends, circles mark phage genomes identified in this work by joining contigs with overlapping termini, the dotted line in the middle of phage033a highlights that this phage was identified at the two termini of a bacterial contig assembly and is missing genes potentially due to an incomplete assembly