Folder Icon Indicates sessions with recordings available.


ANN Modeling of HS-SPME/GC-MS and Sensory Analysis of Potato Clones as a Potential Flavor Prediction Tool during Selective Breeding

Wednesday, August 5, 2015
Napoleon Expo Hall (Sheraton Hotel New Orleans)
Raven A. Bough, BS , Colorado State University, Fort Collins, CO
Most plant breeding programs emphasize increased yield, size, and abiotic or biotic resistance during early selection cycles. However, these foci may inversely affect the production of metabolites that generate produce flavor. Several marketing research studies have indicated that consumers are generally dissatisfied with the flavor of fresh produce and desire increased flavor, suggesting an opportunity to boost consumer appeal through breeding for flavor improvement. Flavor is a complex term that encompasses taste, aroma, and texture. Non-volatile metabolites constitute taste and create basic flavor notes of sweet, sour, bitter, salty, and umami. Aroma consists of volatile metabolites and texture induces characteristic mouthfeel. Because volatile metabolites exhibit more diversified flavor notes than non-volatile metabolites, they are considered to have a more significant effect on the flavor of horticultural produce. Through analysis of named potato cultivars as well as advanced clones grown at Colorado State University’s San Luis Valley Research Center, this study evaluates the utility of an artificial neural network (ANN) to model the relationship between quantitative data and sensory panel analysis of cooked potato tubers. A model that can predict sensory panel response to cooked potato samples based on quantitative data will enable selective breeding for improved flavor, without the arduous task of conducting resource intensive sensory analysis. In this study, the volatile compounds of cooked potato samples will be analyzed using headspace solid-phase microextraction coupled with gas chromatography-mass spectrometry (HS-SPME/GC-MS). A trained panel of 10 to 12 panelists will be used for sensory analysis. An effective ANN model for sensory analysis prediction in potatoes would demonstrate a practical method for flavor evaluation during selective breeding. Selection for flavor improvement in a breeding program will most likely increase consumer appeal of a particular horticultural product, which may effectively facilitate market expansion for that particular crop.