(449a) Connecting Experimental Conditions with Chemical Structure

Reaction
databases provide a great deal of useful information to assist planning of
experiments, but do not provide any interpretation or chemical concepts to
accompany this information.  In this
talk, reactions are labeled with experimental conditions, and network analysis
will show that consistencies within clusters of data points can be leveraged to
interpret underlying concepts that organize the information.  In particular, this analysis will show
why particular choices of solvent are effective in
enabling specific organic reactions (Friedel-Crafts,
aldol addition, Claisen condensation, Diels-Alder,
and Wittig), including variations in solvent choice within each reaction class.  Our strategy will automatically craft
succinct sentences or rules of thumb which describe
the observed solvent selection rules. 
An example of network analysis is shown in Figure 1, where a Friedel-Crafts reaction breaks into clusters
which depend on choice of catalyst, substrate, and stability of the
reaction for/against protic solvent. These
observations, typically belonging to the intuition of a chemist, are derived
directly from the network.

Figure 1.  Friedel-Crafts reactions network.