Postdoctoral Fellow, Monell Chemical Senses Center
Ph.D., Food Science, University of Illinois at Urbana-Champaign
Dr. Joel Mainland
There are millions of known organic molecules in public databases and many orders of magnitude more theoretically possible molecules. How many of these molecules have an odor? What features of the molecule most strongly influence what the molecule will smell like? My research interest lies in using the chemical and physical properties of stimuli to explain and predict human sensory perception. At Monell, I am working to predict odor characteristics of molecules based on their structural characteristics. To build reliable predictive models, we first need reliable datasets. I am applying my background in sensory science and panel training to ensure we collect high quality data that is reproducible and captures small perceptual differences. Combining this perceptual data with a wide array of chemical features, we can build and train models using machine learning to predict both whether a compound is odorous or odorless and what type of odor percept a molecule elicits. These models can help us discover new odorous molecules and identify key molecular features that influence odor perception.
olfaction, psychophysics, machine learning
Mayhew EJ, Neal CH, Lee S-Y, Schmidt SJ. (2017) Glass transition prediction strategies based on the Couchman-Karasz equation in model confectionary systems. J Food Eng. 214, 287-302.
Mayhew EJ, Schmidt SJ, Schlich P, Lee S-Y. (2017) Correlation of consumer perception of stickiness and contributing texture attributes to trained panelist evaluations in a caramel system. Food Qual Prefer. 65, 72-80.
Mayhew EJ, Schmidt SJ, Schlich P, Lee S-Y. (2017) Temporal texture profile and identification of glass transition temperature as an instrumental predictor of stickiness in a caramel system. J Food Sci. 82, 2167-76.