(340ag) Unearthing Enzyme Promiscuity with Cheminformatics to Design Biosynthetic Pathways Towards Novel Biomolecules

Enzyme promiscuity is the ability for enzymes to catalyze a range of side reactions in addition to its main reaction. This is a widely recognized yet still largely unexplored phenomenon in most biological systems, which could open up vast possibilities for bioproduction of valuable chemicals. Novel biosynthetic pathways leading to molecules of interest can be constructed based on enzymes with desired promiscuous reactions, thus allowing for biocatalysis of a wider range of molecules under green conditions.

My research focuses on building cheminformatics knowledge bases and workflows to allow for the computational design and evaluation of novel biosynthetic pathways. First, a set of generalized enzymatic reaction rules capable of describing most enzymatic transformations was curated, which can be applied in the Biochemical Network Integrated Computational Explorer (BNICE) platform to predict the entire space of possible metabolic reactions. These reaction rules were generated by mining public metabolic reaction databases, and extracting a set of reaction SMARTS as well as enzyme sequence information for each type of transformation. Reaction rules were verified to 1) comprehensively cover known reactions across metabolic databases, 2) describe reactions with the maximum level of promiscuity, and 3) represent unique enzymatic transformations.

Based on these rules, a cheminformatics workflow is developed to efficiently predict and prune novel biosynthetic pathways towards molecules of interest, and systematically accelerate the identification of the most promising pathways for biocatalysis. The assessment for each reaction step takes into account factors including thermodynamics, theoretical yield, enzyme availability, and potential deleterious effects of a heterologous enzyme in a biological system. These workflows have been applied to predict biosynthetic pathways towards several biomolecules of commercial interest, currently undergoing experimental validation.