Predicting human olfactory perception by odorant structure and receptor activation profile.

This article focuses on elucidating the relationship between molecular structures, olfactory receptor (OR) activity profiles, and odor perception, specifically examining eugenol, vanillin, and structurally related compounds. The study identified six human ORs primarily activated by these compounds and demonstrated that variations in OR activity profiles correlate with sensory similarity to eugenol, enabling the development of a predictive elastic net regression model (R² = 0.687). Additionally, the authors introduced a novel molecular representation using three-dimensional (3D) shape and pharmacophore fingerprints that consider multiple conformations of odorants, which also predicted perceptual odor similarity with moderate accuracy (R² = 0.514). These findings suggest that integrating OR activity data and advanced 3D molecular profiling can enhance understanding and prediction of olfactory perception within a defined chemical space.