MicrobeWorld

A study published in mBio this week highlights a novel approach to developing new antibiotics for tuberculosis and other infections using high-throughput bioinformatics. Lamichhane et al. generated and genotyped over 5,000 strains of M. tuberculosis and used statistical analysis to find putative essential genes. By cross-listing these essential enzymes with a list of their known products, the team identified molecules that may be necessary for the bacterium. They then selected molecules that were chemically similar to those metabolites – but not identical – the idea being that these compounds can function as competitive inhibitors of essential M. tuberculosis enzymes and kill or interfere with the bacterium’s growth. Sort of a bait-and-switch tactic to pump the bacterium full of a compound that is almost but not quite the thing they need.
Click the "source" link to read more on mBio's blog, mBiosphere...