Title:Support for Eschenmoser's Glyoxylate Scenario

Abstract:A core topic of research in prebiotic chemistry is the search for plausible synthetic routes that connect the building blocks of modern life such as sugars, nucleotides, amino acids, and lipids to "molecular food sources" that have likely been abundant on Early Earth. In a recent contribution, Albert Eschenmoser emphasised the importance of catalytic and autocatalytic cycles in establishing such abiotic synthesis pathways. The accumulation of intermediate products furthermore provides additional catalysts that allow pathways to change over time. We show here that generative models of chemical spaces based on graph grammars make it possible to study such phenomena is a systematic manner. In addition to repro- ducing the key steps of Eschenmoser's hypothesis paper, we discovered previously unexplored potentially autocatalytic pathways from HCN to glyoxylate. A cascading of autocatalytic cycles could efficiently re-route matter, distributed over the combinatorial complex network of HCN hydrolysation chemistry, towards a potential primordial metabolism. The generative approach also has it intrinsic limitations: the unsupervised expansion of the chemical space remains infeasible due to the exponential growth of possible molecules and reactions between them. Here in particular the combinatorial complexity of the HCN polymerisation and hydrolysation networks forms the computational bottleneck. As a consequence, guidance of the computational exploration by chemical experience is indispensable.