The signature genes tool developed for the SEED and implemented at the National Microbial Pathogen Data Resource (NMPDR), www.nmpdr.org, was used to compare the translated genomes of all completely sequenced strains of Streptococcus pyogenes (group A streptococcus or GAS) to define a core genome for this human pathogen. The tool allows the user to select a reference genome to compare with any number of genomes selected in a comparison set. The commonality factor is set to 80% by default but may be reset by the user. For example, the 80% common core of GAS with respect to the strain with the largest genome (MGAS 10750) contains 1,476 proteins that have bidirectional, best BlastP hits (BBH), at an E-value of 1 x 10^-10 or less, in 10 of the 12 available genomes. Increasing the stringency of the analysis to 100% reduces the number of core proteins to 1,368. In addition to determining the proteins in common to a set of genomes, we used the signature genes tool to define a signature set of proteins that distinguishes the strains having the same M-type, e.g. M1 and M12. The information generated by this genome comparison could be used to design a microarray for the simultaneous analysis of the core GAS genome as well as signatures for each sequenced strain or M-type. The bioinformatics analysis reveals interesting consistencies and inconsistencies which generate hypotheses for testing on microarrays. Because protein functions in NMPDR are organized in subsystems, it is possible to infer functional differences imparted by gene signatures. Subsystems annotation is used for metabolic reconstruction, analysis of central machinery and signaling pathways, finding missing genes, integrating regulatory networks, detection of horizontally transferred genes, and prediction of the functions of hypothetical proteins. We show examples on the use of NMPDR applications and the SEED subsystems in understanding the pathogenesis of S. pyogenes, the evolution of its virulent strains, and the study of horizontally transferred toxins. These genome annotation and comparison tools provide unprecedented information about GAS biology and pathogenesis, and can be eventually applied to all sequenced bacterial pathogens.