data-science approach to predicting conditions for C-H oxidation reactions

The research project describes the experimental and computational methods to drive the reactivity of complex molecules towards a more efficient and site-selective chemical transformations focused on unactivated C(sp3)-H bonds in organic synthesis. C-H oxidation has emerged as a power tactic in the synthesis of complex natural products and pharmaceutical agents.

There now exist an array of different methods with different intrinsic selectivity for C-H bonds based on the reagent as well as the steric and electronic properties of the structure. Nevertheless, predicting the conditions for selective C-H oxidations remains challenging on complex substrates.

Importantly, this study will focus on predict selectivity in the transformations to enable aliphatic C-H oxidation as a method for streamlining natural products synthesis in the laboratory.

For this, theoretical investigations will predict the major C-H oxidation site based on intrinsic substrate and reagents reactivity.

These predictions will be carried out  using machine learning approaches to navigate the maze of complex molecules synthesis that involves C-H oxidation processes.

The combination of experimental and computational research will minimize the twists and turns and dead ends that synthetic organic chemists´ asset during the planned reaction.

The research project will be carried out at Caltech in collaboration with experts in the fields of computer science and total synthesis at the NSF C-CAS during 29 months.