A bioinformatics approach to revealing the genetic basis for bacteriophage host specificity

Mycobacteriophages are bacteriophages that infect the genus Mycobacterium, including pathogens such as Mycobacterium tuberculosis and Mycobacterium ulcerans. Full genome sequences of 654 mycobacteriophages are currently available. A mere 20.25% of the 69,581 genes encoded by these phages have at least one known homologue in NCBI, leaving roughly 80% of known mycobacteriophages genes without a predicted function. The host range of 204 mycobacteriophages, initially isolated on Mycobacterium smegmatis strain mc2155, was recently determined on M. tuberculosis and M. smegmatis strains Jucho and MKD8. Three different levels of infectivity were observed: phages that were incapable of infecting the host, phages that were capable of infecting the host, and some that were at a plating efficiency less than one relative to mc2155. The phages that are capable of infecting a host are of particular interest. With so many uncharacterized genes encoded by these phages, we will take a computational approach by performing an association study using the Phamerator software to study the relationship between the complement of protein phamilies in specific genomes and host range of the corresponding phages