UC Berkeley

This dissertation describes our synthesis of the pupukeanane natural products and our application of emerging computer-assisted synthesis techniques to the synthesis of these compounds. In Chapter 1, I review the biosynthesis, biological activities, and previous syntheses of natural products bearing the pupukeanane and allopupukeanane skeleton. Eighteen syntheses are detailed and analyzed according to the key C–C bond forming steps that allow for the synthesis of these complex, tricyclic cores.

Chapter 2, details our efforts to develop a Pd(II)-mediated cascade to form the key bicycles that represent the core of the pupukeanane skeleton and the allopupukeanane skeleton. This approach has culminated in a 10-step formal synthesis of 2-isocyanoallopupukeanane and is the first to provide a strategy toward enantioenriched material. Additionally, a late-stage intermediate bearing the allopupukeanane skeleton is leveraged to allow unified access to the pupukeanane core.

In Chapter 3, an overview of the state-of-the-art in retrosynthetic planning programs is provided. The advent of retrosynthesis occurred nearly 50 years ago, along with the dream of realizing fully automated synthetic planning. In this review, I classify types of planners that are used and the algorithms and computational systems that underlie them. Additionally, I provide analyses of the types of compounds that these programs are equipped to analyze and the limitations in applying them more generally.

Following this review, Chapter 4 details the results of SynthiaTM, an expert-coded synthetic planner that I used to provide synthetic routes to the pupukeanane family. The merits of this program are explored from the lens of academic synthetic chemists and the computer-designed and computer-inspired synthetic strategies are compared and contrasted with previous syntheses of these compounds on which the program has not been explicitly trained.