Tuberculosis (TB), a respiratory disease caused by bacteria in the Mycobacterium tuberculosis (M. tuberculosis) complex, remains one of the top ten causes of death worldwide, with 1.5 million deaths in 2018 alone. One hurdle to eradicating M. tuberculosis is the large portion of its protein-coding genome with unknown function. Assigning function to these genes may lead to new drug targets and a better understanding of mycobacterial physiology and pathogenesis. Here, we describe approaches to characterize genes of unknown function, beginning with analysis of screens from the literature to identify genes likely to be relevant for infection, follow-up in databases that predict functional domains, and subsequent phenotypic analysis of mutants lacking the genes of interest. We utilized this strategy to focus primarily on two genes of unknown function, rv3679 and rv3680, which were predicted to be important for persistence in the mouse model of TB. We show that Rv3679 and Rv3680 interact, and that their ATP-binding domains are required for mediating resistance to elevated levels of nitric oxide and glycerol. Our work uncovered other genes/proteins of unknown function that interact with Rv3679 and/or Rv3680 genetically or physically, underscoring the importance of understanding uncharacterized genes. This includes rv3044 (fecB), which was already known to be important for resistance to front-line antibiotics. We pursued characterization of fecB further, laying a groundwork for future mechanistic studies. Taken together, this work provides experimentally-determined roles for rv3679 and rv3680, with implications for understanding how glycerol may be utilized by the cell outside of the context of a carbon source. This work thus helps reduce the size of the uncharacterized genome, deepens our knowledge of M. tuberculosis physiology, and provides strategies for studying genes of unknown function.