Bayesian models leveraging bioactivity and cytotoxicity information for drug discovery.

Identification of unique leads represents a significant challenge in drug discovery. This hurdle is magnified in neglected diseases such as tuberculosis. We have leveraged public high-throughput screening (HTS) data to experimentally validate a virtual screening approach employing Bayesian models built with bioactivity information (single-event model) as well as bioactivity and cytotoxicity information (dual-event model). We virtually screened a commercial library and experimentally confirmed actives with hit rates exceeding typical HTS results by one to two orders of magnitude. This initial dual-event Bayesian model identified compounds with antitubercular whole-cell activity and low mammalian cell cytotoxicity from a published set of antimalarials. The most potent hit exhibits the in vitro activity and in vitro/in vivo safety profile of a drug lead. These Bayesian models offer significant economies in time and cost to drug discovery.

Importance of confirming data on the in vivo efficacy of novel antibacterial drug regimens against various strains of Mycobacterium tuberculosis.

In preclinical testing of antituberculosis drugs, laboratory-adapted strains of Mycobacterium tuberculosis are usually used both for in vitro and in vivo studies. However, it is unknown whether the heterogeneity of M. tuberculosis stocks used by various laboratories can result in different outcomes in tests of antituberculosis drug regimens in animal infection models. In head-to-head studies, we investigated whether bactericidal efficacy results in BALB/c mice infected by inhalation with the laboratory-adapted strains H37Rv and Erdman differ from each other and from those obtained with clinical tuberculosis strains. Treatment of mice consisted of dual and triple drug combinations of isoniazid (H), rifampin (R), and pyrazinamide (Z). The results showed that not all strains gave the same in vivo efficacy results for the drug combinations tested. Moreover, the ranking of HRZ and RZ efficacy results was not the same for the two H37Rv strains evaluated. The magnitude of this strain difference also varied between experiments, emphasizing the risk of drawing firm conclusions for human trials based on single animal studies. The results also confirmed that the antagonism seen within the standard HRZ regimen by some investigators appears to be an M. tuberculosis strain-specific phenomenon. In conclusion, the specific identity of M. tuberculosis strain used was found to be an important variable that can change the apparent outcome of in vivo efficacy studies in mice. We highly recommend confirmation of efficacy results in late preclinical testing against a different M. tuberculosis strain than the one used in the initial mouse efficacy study, thereby increasing confidence to advance potent drug regimens to clinical trials.