PRISM: a computational resource for genomic natural product discovery

Microbial life is an invaluable source of pharmaceutically relevant small molecules, but while genomic studies indicate nearly 90% remain undiscovered, industrial programs are increasingly frustrated by high rediscovery rates. Connecting biosynthetic gene clusters to new chemical scaffolds is critical to enabling continued natural product discovery in a post-genomic era. Elucidation of the modular strategies employed by nature in the synthesis of nonribosomal peptides and polyketides has defined an opportunity to leverage a wealth of genomic data in search of new chemistry. PRISM is a resource for rapid genomic prediction of microbial secondary metabolomes and informatic identification of unknown natural products. PRISM’s algorithm creates a combinatorial library of chemical structures representing the organism’s secondary metabolome, and leverages databases of all known natural products and biosynthetic gene clusters to dereplicate known metabolites, thereby identifying unknown natural products within genetic data. As a rapid tool to selectively identify novel chemistries by chemo- and bioinformatic analysis with a web-based user interface accessible to non-specialists, PRISM represents an important new tool for the discovery of chemical architectures.