Project 454957

Microorganisms residing in the gut significantly influence human health, such that sudden changes in the structure or diversity of gut microbiota has been associated with several metabolic and neurodegenerative disorders. Although the mechanistic know-how of this regulation by these miniscule organisms in the complex host environment, remains largely unknown; recent advances in the fields of gut metagenomics and metabolomics have established the interaction of gut associated small molecules with their host. These natural products produced by gut bacteria have the potential to be used as therapeutics, much like most of the drugs derived from soil or marine environments, during the early 20th century. This project aims to explore the "gold mine" of gut microbiota derived molecules and to discover their therapeutic potential, by employing an engineered set of E. coli strains tagged with a fluorescent reporter. Our group has recently established the potential of combining promoter-reporter libraries with an in-house developed hardware called the Promoter Fluorescent Imaging (PFI) box, to acquire genome-wide transcriptional responses to chemical stresses. Since, E. coli is the pioneer organism of the human gut, I aim to exploit this approach and build a training set in the presence of a wide variety of compounds that include all modern-day drugs like anticancers, antipsychotics, antidepressants etc. Reporter arrays when bombarded with these drugs would elicit unique transcriptional responses, which would be used to train a neural network and this information will then be embedded into a deep learning model. This neural network will have the capacity to identify biological activity matching any of the drugs in the training set and thus could provide strong hypotheses for the therapeutic potential of any compound tested. Herein, I propose to focus on purified metabolites derived from cultures of gut-associated bacteria and to predict their therapeutic potential.