A focused collection of organic synthesis reactions for computer-based molecule construction is presented. It is inspired by real-world chemistry and has been compiled in close collaboration with medicinal chemists to achieve high practical relevance. Virtual molecules assembled from existing starting material connected by these reactions are supposed to have an enhanced chance to be amenable to real chemical synthesis. About 50% of the reactions in the dataset are ring-forming reactions, which fosters the assembly of novel ring systems and innovative chemotypes. A comparison with a recent survey of the reactions used in early drug discovery revealed considerable overlaps with the collection presented here. The dataset is available encoded as computer-readable Reaction SMARTS expressions from the Supporting Information presented for this paper.

The use of flow photochemistry and its apparent superiority over batch has been reported by a number of groups in recent years. To rigorously determine whether flow does indeed have an advantage over batch, a broad range of synthetic photochemical transformations were optimized in both reactor modes and their yields and productivities compared. Surprisingly, yields were essentially identical in all comparative cases. Even more revealing was the observation that the productivity of flow reactors varied very little to that of their batch counterparts when the key reaction parameters were matched. Those with a single layer of fluorinated ethylene propylene (FEP) had an average productivity 20% lower than that of batch, whereas three-layer reactors were 20% more productive. Finally, the utility of flow chemistry was demonstrated in the scale-up of the ring-opening reaction of a potentially explosive [1.1.1] propellane with butane-2,3-dione.

In the search for new bioactive compounds, there is a trend toward increasingly complex compound libraries aiming to target the demanding targets of the future. In contrast, medicinal chemistry and traditional library design rely mainly on a small set of highly established and robust reactions. Here, we probe a set of 58 such reactions for their ability to sample the chemical space of known bioactive molecules, and the potential to create new scaffolds. Combined with ~26,000 common available building blocks, the reactions retrieve around 9% of a scaffold-diverse set of compounds active on human target proteins covering all major pharmaceutical target classes. Almost 80% of generated scaffolds from virtual one-step synthesis products are not present in a large set of known bioactive molecules for human targets, indicating potential for new discoveries. The results suggest that established synthesis resources are well suited to cover the known bioactivity-relevant chemical space and that there are plenty of unexplored regions accessible by these reactions, possibly providing valuable "low-hanging fruit" for hit discovery.

No abstract

No abstract