Increased resolution of African swine fever virus genome patterns based on profile HMMs of protein domains.

African swine fever virus (ASFV), belonging to the Asfarviridae family, was originally described in Africa almost 100 years ago and is now spreading uncontrolled across Europe and Asia and threatening to destroy the domestic pork industry. Neither effective antiviral drugs nor protective vaccines are currently available. Efforts to understand the basis for viral pathogenicity and the development of attenuated potential vaccine strains are complicated by the large and complex nature of the ASFV genome. We report here a novel alignment-free method of documenting viral diversity based on profile hidden Markov model domains on a genome scale. The method can be used to infer genomic relationships independent of genome alignments and also reveal ASFV genome sequence differences that determine the presence and characteristics of functional protein domains in the virus. We show that the method can quickly identify differences and shared patterns between virulent and attenuated ASFV strains and will be a useful tool for developing much-needed vaccines and antiviral agents to help control this virus. The tool is rapid to run and easy to implement, readily available as a simple Docker image.