“Machine-learning-based odor predictive” tech for endless tailored fragrance impressions

05 Sep 2022 --- Machine learning models and mass spectrum data have made it possible to tailor odors to create an intended impression. The researchers at the Tokyo Institute of Technology have developed a creative solution that predicts molecular features of scents, corresponding to a desired impression. This method has potential applications in the food, health, beauty and wellness industries.

Olfactory receptors in olfactory nerve cells are how humans can detect smells. The olfactory impressions of odorants on nerve cells are tied to the odor’s molecular features and physicochemical properties.

“We used a machine-learning-based odor predictive model that we had previously developed to obtain the odor impression,” explains Professor Takamichi Nakamoto, the lead researcher at the Tokyo Institute of Technology.

“Then we predicted the mass spectrum from odor impression inversely based on the previously developed forward model.” Scientists now make it possible to tailor odors to create an intended odor impression. 

The new method is touted to provide highly accurate predictions of the physicochemical properties of odor mixtures and the mixing ratios required to prepare them. Therefore, “opening the door to endless tailor-made fragrances.” 

Meeting the desired impression
The mass spectra of odor mixtures were obtained by a linear combination of the mass spectra of single components. According to the researchers, this method can quickly prepare for the ranges of odor mixtures and the required mixing ratio, which is essential when preparing new odors. 

“For example, we show which molecules give the mass spectrum of apple flavor with enhanced ‘fruit’ and ‘sweet’ impressions. Our analysis shows that combinations of either 59 or 60 molecules give the same mass spectrum as the one obtained from the specified odor impression. With this information, and the correct mixing ratio needed for a certain impression, we could theoretically prepare the desired scent,” highlights Nakamoto.

“We proposed mathematical models and algorithms for calculating physicochemical features, i.e., the mass spectrum that realizes an intended odor impression.”The scientists proposed mathematical models and algorithms for calculating physicochemical features. 

Workings of olfactory receptors
The scientists highlight that humans perceive thousands of odorants through approximately 400 types of olfactory receptors. One kind of sensory receptor is expressed on an olfactory nerve cell. The response patterns of the olfactory receptor to an odorant are the same and have the same odor selectivity. 

“Then, the brain processes the response patterns of olfactory nerve cells that are differently excited by odorants having various types of physicochemical information,” the scientists share.

They note that as sensory tests require time and labor and attempts have been made to predict sensory test data from physicochemical parameters or sensing data. “Many studies use physicochemical parameters to predict odor perception, but they do not apply to odor mixtures.”

Edited by Venya Patel