In Silico Exploration of Metabolite-Derived Soft Materials Using a Chemical Reaction Network

Future soft materials and polymer chemistries will require innovative nonpetroleum sourcing pathways. While leveraging microbial metabolites derived from biological feedstocks possesses high potential in many avenues of chemical development, the applicability of this paradigm to the specifics of soft materials chemistry is unclear. Here, we construct a chemical reaction network based on databases of common microbial metabolites and the USPTO reaction set to examine what is possible in the chemical space of metabolite-derived chemistries of relevance to soft materials. We observe that the accessible chemical space of our chemical reaction network possesses strong microbe-specific chemical diversity and that this space saturates rapidly within three synthetic steps applied to the original microbial metabolites. Importantly, we show that the chemical space accessible from metabolite precursors possesses significant overlap with existing petrochemical building blocks, known and proposed synthetically feasible polymer monomers, and the chemical space of common organic semiconductors and redox active materials. The biases induced by the metabolite and reaction databases that parametrize our reaction network are analyzed as a function of chemical functional groups, and pathways toward broader sets of chemistries and reactions are outlined. This work introduces a computational framework for soft materials discovery with the potential to accelerate the identification of soft materials relevant to metabolic engineering targets and nonpetroleum sourcing pathways for existing soft materials.