Jerome Golebiowski - Institut de Chimie de Nice, UMR CNRS 7272 - Keywords: smell, taste, numerical models, molecular modeling

Jerome Golebiowski - Institut de Chimie de Nice, UMR CNRS 7272 - Keywords: smell, taste, numerical models, molecular modeling

Contribution title: Cracking the code of chemosensory perception using computational tools

Chemical senses such as smell and taste play vital roles in our daily life. In addition to basic odor and taste perception, they are also strongly associated with our psychophysiological responses including emotions and memories. Upon interactions of external chemicals with our chemosensory receptors (e.g. odorant receptors in the nose or taste receptors on the tongue), extremely complex signaling processes are translated in the brain by billions of interconnected neurons. Decoding chemosensory perception will lead to wide applications in the research fields of flavor, fragrance, medicine and related products for our well being. We attempt to address this highly multidisciplinary challenge by connecting machine learning algorithms and molecular modeling with in vitro and ex vivo sensory analyses and neurophysiology experiments on human individuals.

In practice, machine learning enables analysis and prediction of the structure-activity relationship of large number of compounds. Molecular modeling studies atomistic-level interactions of these compounds with their receptors. In vitro / in vivo assays assess the activity of these compounds on given receptors, or human panelists, providing feedbacks to optimize the computational algorithms and protocols. Physiology experiments on humans provide direct readouts of the compounds’ psychophysiological effects. With experts in each of these fields on board, we will gain profound understanding of our chemosensory perception, which will allow the rational design of new compounds with desired properties to be tested.
We'll hopefully soon be able to say "Draw me an odor, draw me a taste, draw me an emotion"!