Daniel Graham

My research interests lie in the application of information theory to thermodynamic systems, including molecules, an endeavor at the intersection of thermodynamics, probability and statistics, and computer science. While information theory was developed originally to understand—and thus improve—communication systems, it provides yet another powerful tool by which the relation between molecular structure and function can be comprehended.   My research is a search for new ways to predict molecular activity in advance of experiment, and to understand thermodynamic algorithms at the micro- and macroscopic scales.

My approach to the microscopic scale concentrates on the electronic messages transmitted by molecules during Brownian motion and collisions.  These messages are represented pictorially by structure diagrams combined with a simple random walk model.  Thermochemical principles are able to address the energetics of the messages; probability fundamentals enable the message information to be quantified and correlated.   Lastly, conceptuals drawn from computer science establish connections between the messages and their chemical effects.  Overall, the results are new quantitative structure/activity relations (QSARs) for organic compounds.  These are of significance to several disciplines including biological, medicinal, and industrial organic chemistry.