26 March 2024 in Nature Communications
The perception and appreciation of food flavor depends on many interacting chemical compounds and external factors, and therefore proves challenging to understand and predict. Here, we combine extensive chemical and sensory analyses of 250 different beers to train machine learning models that allow predicting flavor and consumer appreciation. For each beer, we measure over 200 chemical properties, perform quantitative descriptive sensory analysis with a trained tasting panel and map data from over 180,000 consumer reviews to train 10 different machine learning models. The best-performing algorithm, Gradient Boosting, yields models that significantly outperform predictions based on conventional statistics and accurately predict complex food features and consumer appreciation from chemical profiles. Model dissection allows identifying specific and unexpected compounds as drivers of beer flavor and appreciation. Adding these compounds results in variants of commercial alcoholic and non-alcoholic beers with improved consumer appreciation. Together, our study reveals how big data and machine learning uncover complex links between food chemistry, flavor and consumer perception, and lays the foundation to develop novel, tailored foods with superior flavors.
24 October 2016 in PLOS ONE
Saccharomyces cerevisiae is routinely used yeast in food fermentations because it combines several key traits, including fermentation efficiency and production of desirable flavors. However, the dominance of S. cerevisiae in industrial fermentations limits the diversity in the aroma profiles of the end products. Hence, there is a growing interest in non-conventional yeast strains that can help generate the diversity and complexity desired in today’s diversified and consumer-driven markets. Here, we selected a set of non-conventional yeast strains to examine their potential for bread fermentation. Here, we tested ten non-conventional yeasts for bread fermentation, including two Saccharomyces species that are not currently used in bread making and 8 non-Saccharomyces strains. The results show that Torulaspora delbrueckii and Saccharomyces bayanus combine satisfactory dough fermentation with an interesting flavor profile. Sensory analysis and HS-SPME-GC-MS analysis confirmed that these strains produce aroma profiles that are very different from that produced by a commercial bakery strain. Moreover, bread produced with these yeasts was preferred by a majority of a trained sensory panel. These results demonstrate the potential of T. delbrueckii and S. bayanus as alternative yeasts for bread dough leavening, and provide a general experimental framework for the evaluation of more yeasts and bacteria.
1 December 2018 in Food Microbiology
27 June 2017 in Functional Ecology
1 February 2021 in International Journal of Food Microbiology