Mapping the Chemical Space of Synthetic and Natural Product Chelators

Secondary metabolite natural products are recognized as having high structural diversity, with the associated property diversity resulting in their prevalent clinical use, compared to lower-diversity synthetic agents. It follows that subclasses of secondary metabolite natural products with useful applications, such as metal chelators (including siderophores), would have higher structural diversity compared to synthetic chelator counterparts, however, there are no quantitative measures to support this assumption. Chelators have wide use in applications across metal recovery and radiometal delivery, with growth in these sectors leading to expanded efforts to source new chelators. This motivated a quantitative analysis of the structural diversity of natural product and synthetic chelators. This work produced a data set of chelators (n = 35,544) extracted from databases and literature, and annotated members as either natural product chelators (NC) (n = 2784) or synthetic chelators (SC) (n = 32,760). Molecular descriptors relevant to the structural features of chelators were selected and the property representations used to evaluate trends in the NC and SC data sets and to generate a map of chemical space using principal component analysis (PCA) followed by kernel density estimates (KDE) to measure population densities. While there was some overlap in the regions of chemical space occupied by the NC and SC data sets, compared to the SC data set, the NC data set occupied a larger area (A _[NC] = 182, A _[SC] = 114) with members more diffuse, as quantified by an increased average KDE value (d _av[NC] = 7.61, d _av[SC] = 4.58). These quantitative data confirm natural product chelators are more structurally diverse than synthetic chelators. Exemplars from the NC and SC data sets that occupied the chemical space extrema were pyoverdine and substituted porphyrin compounds, respectively. Natural product chelators comprise a set of structural features that reflect a structural diversity not represented in synthetic chelators. This could inform molecular design toward chelators with more diverse properties to expand applications.