Asset Details
Do you wish to reserve the book?
Predictive Models of Odor Contribution and Thresholds for Volatiles in Identification of Novel Crop Aroma Compounds
Hey, we have placed the reservation for you!
By the way, why not check out events that you can attend while you pick your title.
Oops! Something went wrong.
Looks like we were not able to place the reservation. Kindly try again later.
Are you sure you want to remove the book from the shelf?
Predictive Models of Odor Contribution and Thresholds for Volatiles in Identification of Novel Crop Aroma Compounds
Do you wish to request the book?
Predictive Models of Odor Contribution and Thresholds for Volatiles in Identification of Novel Crop Aroma Compounds
Please be aware that the book you have requested cannot be checked out. If you would like to checkout this book, you can reserve another copy
We have requested the book for you!
Your request is successful and it will be processed during the Library working hours. Please check the status of your request in My Requests.
Oops! Something went wrong.
Looks like we were not able to place your request. Kindly try again later.
Journal Article
Predictive Models of Odor Contribution and Thresholds for Volatiles in Identification of Novel Crop Aroma Compounds
2025
Overview
Background/Objectives: Aroma is a key determinant of crop quality and consumer acceptance, and aroma contribution and odor threshold are critical attributes for the identification of aroma compounds. Because the experimental determination of aroma contribution and odor thresholds is time-consuming and complex, most volatiles lack contribution and/or threshold data. Methods: We compiled odor thresholds for 716 volatile compounds and 31,459 aroma contribution records, and trained machine-learning models that took molecular fingerprints and physicochemical descriptors (e.g., molecular weight, logP, TPSA) as inputs to predict aroma contribution and odor threshold. We evaluated multiple fingerprint–model combinations, optimized hyperparameters via 5-fold cross-validation on the training set, and assessed the best models on a held-out validation set. Results: The ECFP6–GBDT combination performed best for predicting aroma contribution (macro-F1 = 0.732; weighted-F1 = 0.912). The ECFP4–GBDT model performed best for predicting odor thresholds (R2 = 0.94; RMSE = 0.44). Applying the models to volatiles in passion fruit juice identified 2-phenylethyl acetate as a potential new contributor to passion fruit aroma, whereas menthyl acetate likely exerted a negative influence; both findings were confirmed by serial dilution and sensory evaluation. The developed models provided both a GUI and a CLI, were easy to use, and supported straightforward upgrades by retraining with user-provided data. Conclusions: This work provided a methodological foundation for identifying crop aroma compounds and supported the genetic improvement of aroma traits.
