Prophage genomics reveals patterns in phage genome organization and replication

Han Suh Kang; Katelyn McNair; Daniel A. Cuevas; Barbara A. Bailey; Anca M. Segall; Robert A. Edwards

Identification of prophages in bacterial genomes is complicated by the problem that many phage genes do not share sequence homology with other proteins. In some cases a newly discovered phage does not carry any genes that share sequence similarity to any other gene in the available databases, highlighting how little we know about these viruses (13) . The underrepresentation of wellannotated phage genes in available databases makes it difficult to assign functions to newly sequenced genomes. This is exacerbated as genome sequencing increases at a near-exponential rate, requiring reliance on automated annotation software that is dependent on primary sequence similarity to assign functions to new genes (14, 15) . Most prophage identification algorithms rely on sequence similarity to known phage genes to identify prophage regions (16) (17) (18) (19) (20) . This leads to a paradoxical problem of not being able to identify a prophage that does not contain any regions homologous to known phage proteins. Our prophage-finding program, PhiSpy, uses a novel algorithm that is capable of identifying a prophage even if there is no similarity to any of the proteins in the databases by analyzing a suite of genomic characteristics that separate phage-encoding regions from the bacterial core of the genome (21) .