Abstract
In synthesis planning, the goal is to synthesize a target molecule from available starting materials, possibly optimizing costs such as price or environmental impact of the process. Current algorithmic approaches to synthesis planning are usually based on selecting a bond set and finding a single good plan among those induced by it. We demonstrate that synthesis planning can be phrased as a combinatorial optimization problem on hypergraphs by modeling individual synthesis plans as directed hyperpaths embedded in a hypergraph of reactions (HoR) representing the chemistry of interest. As a consequence, a polynomial time algorithm to find the K shortest hyperpaths can be used to compute the K best synthesis plans for a given target molecule. Having K good plans to choose from has many benefits: it makes the synthesis planning process much more robust when in later stages adding further chemical detail, it allows one to combine several notions of cost, and it provides a way to deal with imprecise yield estimates. A bond set gives rise to a HoR in a natural way. However, our modeling is not restricted to bond set based approaches—any set of known reactions and starting materials can be used to define a HoR. We also discuss classical quality measures for synthesis plans, such as overall yield and convergency, and demonstrate that convergency has a built-in inconsistency which could render its use in synthesis planning questionable. Decalin is used as an illustrative example of the use and implications of our results.
Highlights
Synthesis planning is a core problem in chemistry, first treated as computational problem by Corey [1] in the late sixties
We have demonstrated that a core part of chemical synthesis planning can be phrased as a combinatorial optimization problem on hypergraphs by modeling individual synthesis plans as directed hyperpaths embedded in a hypergraph of reactions (HoR) representing the chemistry of interest
An immediate consequence is that the K best synthesis plans for a given target can be computed in polynomial time [20] for a number of quality measures of practical importance, and for any measure which can be expressed as an additive weight function
Summary
Synthesis planning is a core problem in chemistry, first treated as computational problem by Corey [1] in the late sixties. Synthesis planning is still regarded as somewhat of an art form, attempts have been made over several decades at applying formal approaches and computational methods [1,2,3,4,5,6,7,8,9,10,11,12,13,14,15]. An early contribution towards an automated approach for synthesis planning is retrosynthetic analysis It was introduced by Corey [1, 12, 13] in 1969 as part of a Fagerberg et al J Cheminform (2018) 10:19
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
