In 2006, I worked at Pasteur Institute (Sylvain Brisse groups) on characterization of special smooth strains isolated in Africa1 which seems to be related to M.tuberculosis (known to have a rough phenotype). Using 16 housekeeping genes, we showed that those strains belong to the same species of M. tuberculosis showing an higher diversity than previously observed.
Following this work, in 2013, the Pasteur institute finally published2 a set of genome of those smooth strains concluding that M. tuberculosis emerged from an ancestral pool of mycobacteria by gain of persistence and virulence mechanisms.