Unser 10. Leibniz Plant Biochemistry Symposium am 7. und 8. Mai war ein großer Erfolg. Thematisch ging es in diesem Jahr um neue Methoden und Forschungsansätze der Naturstoffchemie. Die exzellenten Vorträge über Wirkstoffe…
Omanische Heilpflanze im Fokus der Phytochemie IPB-Wissenschaftler und Partner aus Dhofar haben jüngst die omanische Heilpflanze Terminalia dhofarica unter die phytochemische Lupe genommen. Die Pflanze ist reich an…
Geschmack ist vorhersagbar: Mit FlavorMiner. FlavorMiner heißt das Tool, das IPB-Chemiker und Partner aus Kolumbien jüngst entwickelt haben. Das Programm kann, basierend auf maschinellem Lernen (KI), anhand der…
Herrera-Rocha, F.; Fernández-Niño, M.; Duitama, J.; Cala, M. P.; Chica, M. J.; Wessjohann, L. A.; Davari, M. D.; Barrios, A. F. G.;FlavorMiner: a machine learning platform for extracting molecular flavor profiles from structural dataJ. Cheminform.16140(2024)DOI: 10.1186/s13321-024-00935-9
Flavor is the main factor driving consumers acceptance of food products. However, tracking the biochemistry of flavor is a formidable challenge due to the complexity of food composition. Current methodologies for linking individual molecules to flavor in foods and beverages are expensive and time-consuming. Predictive models based on machine learning (ML) are emerging as an alternative to speed up this process. Nonetheless, the optimal approach to predict flavor features of molecules remains elusive. In this work we present FlavorMiner, an ML-based multilabel flavor predictor. FlavorMiner seamlessly integrates different combinations of algorithms and mathematical representations, augmented with class balance strategies to address the inherent class of the input dataset. Notably, Random Forest and K-Nearest Neighbors combined with Extended Connectivity Fingerprint and RDKit molecular descriptors consistently outperform other combinations in most cases. Resampling strategies surpass weight balance methods in mitigating bias associated with class imbalance. FlavorMiner exhibits remarkable accuracy, with an average ROC AUC score of 0.88. This algorithm was used to analyze cocoa metabolomics data, unveiling its profound potential to help extract valuable insights from intricate food metabolomics data. FlavorMiner can be used for flavor mining in any food product, drawing from a diverse training dataset that spans over 934 distinct food products.Scientific Contribution FlavorMiner is an advanced machine learning (ML)-based tool designed to predict molecular flavor features with high accuracy and efficiency, addressing the complexity of food metabolomics. By leveraging robust algorithmic combinations paired with mathematical representations FlavorMiner achieves high predictive performance. Applied to cocoa metabolomics, FlavorMiner demonstrated its capacity to extract meaningful insights, showcasing its versatility for flavor analysis across diverse food products. This study underscores the transformative potential of ML in accelerating flavor biochemistry research, offering a scalable solution for the food and beverage industry.
Publikation
Herrera-Rocha, F.; León-Inga, A. M.; Aguirre Mejía, J. L.; Rodríguez-López, C. M.; Chica, M. J.; Wessjohann, L. A.; González Barrios, A. F.; Cala, M. P.; Fernández-Niño, M.;Bioactive and flavor compounds in cocoa liquor and their traceability over the major steps of cocoa post-harvesting processesFood Chem.435137529(2024)DOI: 10.1016/j.foodchem.2023.137529
The production of fine-flavor cocoa represents a promising avenue to enhance socioeconomic development in Colombia and Latin America. Premium chocolate is obtained through a post-harvesting process, which relies on semi-standardized techniques. The change in the metabolic profile during cocoa processing considerably impacts flavor and nutraceutical properties of the final product. Understanding this impact considering both volatiles and non-volatile compounds is crucial for process and product re-engineering of cocoa post-harvesting. Consequently, this work studied the metabolic composition of cocoa liquor by untargeted metabolomics and lipidomics. This approach offered a comprehensive view of cocoa biochemistry, considering compounds associated with bioactivity and flavor in cocoa liquor. Their variations were traced back over the cocoa processing (i.e., drying, and roasting), highlighting their impact on flavor development and the nutraceutical properties. These results represent the basis for future studies aimed to re-engineer cocoa post-harvesting considering the variation of key flavor and bioactive compounds over processing.
Publikation
Herrera-Rocha, F.; Cala, M. P.; León-Inga, A. M.; Aguirre Mejía, J. L.; Rodríguez-López, C. M.; Florez, S. L.; Chica, M. J.; Olarte, H. H.; Duitama, J.; González Barrios, A. F.; Fernández-Niño, M.;Lipidomic profiling of bioactive lipids during spontaneous fermentations of fine-flavor cocoaFood Chem.397133845(2022)DOI: 10.1016/j.foodchem.2022.133845
The impact of cocoa lipid content on chocolate quality has been extensively described. Nevertheless, few studies have elucidated the cocoa lipid composition and their bioactive properties, focusing only on specific lipids. In the present study the lipidome of fine-flavor cocoa fermentation was analyzed using LC-MS-QTOF and a Machine Learning model to assess potential bioactivity was developed. Our results revealed that the cocoa lipidome, comprised mainly of fatty acyls and glycerophospholipids, remains stable during fine-flavor cocoa fermentations. Also, several Machine Learning algorithms were trained to explore potential biological activity among the identified lipids. We found that K-Nearest Neighbors had the best performance. This model was used to classify the identified lipids as bioactive or non-bioactive, nominating 28 molecules as potential bioactive lipids. None of these compounds have been previously reported as bioactive. Our work is the first untargeted lipidomic study and systematic effort to investigate potential bioactivity in fine-flavor cocoa lipids.