The advent of whole-genome sequencing of bacteria and advances in bioinformatics have revolutionized the study of bacterial pathogenesis, enabling the targeting of possible vaccine candidates starting from genomic information. Nowadays, the availability of hundreds of bacterial genomes enables identification of the genetic differences across several genomes from the same species. The unexpected degree of intra-species diversity suggests that a single genome sequence is not entirely representative and does not offer a complete picture of the genetic variability of a species. The practical consequence is that, in many cases, a universal vaccine is possible only by including a combination of antigens and this combination must take into account the pathogen population structure.