Machine Learning Toolset for Reaction Condition Determination

Professor Scott Denmark from the University of Illinois has developed a machine learning (ML) toolset for reaction discovery. This ML toolset can be used to predict reaction outcomes, including yield, and reaction conditions for various silicon cross-coupling reactions that are included within ~1400 reaction database. But, more importantly, this toolset can identify effective reaction conditions for novel cross coupling reactions. This ML toolset meets a long felt need in the area of reaction discovery by providing a fast and easy to use chemical informatics system that provides accurate results, that in conjunction with user expertise minimizesthe need to extrapolate into unrepresented chemical space.