Before the availability of high-resolution genotyping tools in 1990s, there was a prevailing dogma of little genomic sequence diversity in Mycobacterium tuberculosis. Due to the low levels of genetic variation, it was assumed that M. tuberculosis exhibit very little phenotypic variation in immunologic and virulence factors. The fingerprinting method based on restriction fragment length polymorphisms (RFLP) of IS6110 insertion sequences had unveiled the underestimation of the sequence variation in M. tuberculosis and the importance of strain-to-strain variation for understanding pathogenesis, immune mechanisms, bacterial evolution, and host adaptation. This method became a gold standard for strain differentiation in the molecular epidemiological study. It had lead to a profusion of studies in molecular epidemiology such as the detection of unsuspected transmission, the estimation of the extent of recent transmission, the identification of laboratory cross-contamination, the identification of outbreaks, and distinction between reinfection and relapse. This, in 1990s, is the opening of the molecular epidemiology of tuberculosis. After the completion of genome project of the M. tuberculosis laboratory strain H37Rv, some of the clinical isolates were completely sequenced. This prompted the in silico genome comparison and identified various genomic markers which can give a unifying framework for both epidemiology and evolutionary analysis of M. tuberculosis population. Of them, variable numbers of tandem repeats (VNTR) was found as the most promising PCR-based method which can provide adequate discrimination of M. tuberculosis strains in many cases, including the estimation of M. tuberculosis transmission and the identification of genetic lineages. PCR-based VNTR analysis is easy, rapid, and highly specific and can generate portable digit-based data, unlike the analog information obtained from IS6110 RFLP which is labor intensive. In this regards, investigators can easily compare the genotypic data of independent studies between different laboratories. With the advantages, VNTR surpassed IS6110 RFLP and became the first line genotyping method in molecular epidemiology. One of the most attractive potentials on this method is its applicability for establishment of the database of M. tuberculosis genotype which covers not only local area but also world wide scale. This would open the door to "in silico epidemiology" which brings a breakthrough on the current TB control program. The optimization and standardization of the combination of VNTR loci for strain genotyping is the only but hard issue for the development of global database system. Road to the global Mtb genotype database is hard, but we believe, "Yes, We Can!". Another attractive potential of VNTR is its use for phylogenetic analysis, although more intensive research on this with using comprehensive marker sets, such as large sequence polymorphisms and single-nucleotide polymorphisms are required. Again, with the advantages of VNTR analysis, i.e., easy, rapid, specific, and digit-based data, VNTR became the first line method in molecular epidemiology. Molecular epidemiology of tuberculosis is expanding its research field from the investigation of TB transmission to more basic science such as evolution and phylogeographic distribution. In this symposium, we have invited four opinion leaders in molecular epidemiology of TB in Japan who are talking about each title as followed. 1. Establishment of the standard VNTR analysis systems for Tuberculosis (TB) and preparation of databases for TB genotyping: Shinji MAEDA and Yoshiro MURASE (Department of Mycobacterium Reference and Research, Research Institute of Tuberculosis, JATA). We have already reported the JATA (12)-VNTR system for TB genotyping in Japan. However, by comparison of cluster formation rate, the discrimination power of JATA (12)-VNTR was lower than that of IS6110 RFLP analysis. Therefore, we improved the JATA (12)-VNTR system for developing discrimination power. By addition of 3 loci (ETR-A, VNTR-1982 and VNTR-2163 a) to JATA (12)-VNTR, we established new JATA (15)-VNTR. We found that the discrimination power of JATA (15)-VNTR was almost the same as that of RFLP analysis. 2. Molecular epidemiology of Mycobacterium tuberculosis reviewed by molecular epidemiology of other pathogenic bacteria: Eiji YOKOYAMA (Division of Bacteriology, Chiba Prefectural Institute of Public Health). Molecular epidemiology of M. tuberculosis should be progressed to two goals. First is the short-term goal that intends to elucidate the unapparent route of transmission of the organism. Second is the long-term goal that intends to ascertain the phylogeny of the organism. The combination of VNTR loci should be changed according to the goals of molecular epidemiology of the organism. 3. Progress of the research in molecular epidemiology of Mycobacterium tuberculosis: Tomotada IWAMOTO (Department of Microbiology, Kobe Institute of Health). In the past decade, molecular epidemiology of tuberculosis brought significant insights into the transmission of tuberculosis, genetic diversity of M. tuberculosis, population structure and geographical distribution of M. tuberculosis, etc. In the advanced stage of the molecular epidemiological study, we expect to change the current geno-typing based molecular epidemiology to whole genome-typing based molecular epidemiology on the basis of the rapid innovation of next-generation sequencing technology. 4. Clinical application of molecular epidemiology of tuberculosis: Tomoshige MATSUMOTO (Department of Clinical Research and Development, Osaka Prefectural Medical Center for Respiratory and Allergic Diseases). The molecular epidemiology can be applied in clinical practice. We showed some examples about usefulness of the clinical application of molecular epidemiology, especially using variable number of tandem repeats (VNTR) analysis. One example we showed: using VNTR, we can know whether two tuberculosis bacilli which developed from the patients, who have close contact, are the same or not in a few days; Especially, when one patient suffers from multidrug-resistant (MDR) strain of or extensively drug resistant (XDR) of tuberculosis, we can easily know whether the other suffers from MDR/XDR tuberculosis or not. The other example we showed: we can know relapse, reinfection, or laboratory contamination by using VNTR in a few days when a patient shows bacteriological relapse during the treatment. By introducing VNTR to clinical practice, we can diminish days of inappropriate hospitalization. Because VNTR data are numerical, we can easily construct VNTR database, compare data, and survey emergence of MDR/XDR-tuberculosis.