From components to communities: bringing network science to clustering for genomic epidemiology