The complicated and dynamic nature of microbiome investigation can benefit from systems-biology approaches including in silico genome-wide computation and in vivo genome-scale experimental measurement. In this talk, three systems toolsets will be introduced, which can be applied to study individual bacterium or microbiome: pan-genome analysis, analysis of transcriptional regulatory network, and genome-scale metabolic modeling. Pan-genome analysis was used to study genomic features of newly isolated M. avium subsp. Paratuberculosis strains, which can cause Johne’s disease in cows. Integration of machine-learning based transcription factor prediction and experimental validation with ChIP-exo enabled investigation of less-studied fraction of transcriptional regulation in bacteria. Lastly, genome-scale metabolic modeling provided explanatory and predictive capabilities for understanding bacterial metabolic at the genome scale, which can be further used in metabolic interactions in bacterial community.