Bacteriophages are the most predominant and genetically diverse biological entities on Earth. They are bacterial viruses which encode numerous proteins with potential antibacterial activity. However, most bacteriophage-encoded proteins have no assigned function, hindering the discovery of novel antibacterial agents. In particular, there has been a growing interest in exploring recombinant bacteriophage depolymerases from the fundamental standpoint, but mostly for biotechnological applications to control bacterial pathogens. Due to the lack of efficient identification tools, we developedPhageDPO, the first developed tool that predicts depolymerases in bacteriophage genomes using machine learning methods.Availability and implementationPhageDPOwas integrated into a Galaxy framework available online at: bit.ly/phagedpo.