Characterization of Volatile Compounds in New Cider Apple Genotypes Using Multivariate Analysis
Gas chromatography combined with solid-phase microextraction has been used for the identification of the aromatic profiles of new cider apple genotypes, and a chemometric characterization of these new cider apple genotypes has been carried out using exploratory and modelling techniques. Three breedi...
| Autores: | , , , , , , |
|---|---|
| Tipo de recurso: | artículo |
| Fecha de publicación: | 2016 |
| País: | España |
| Institución: | Universidad de Oviedo (UNIOVI) |
| Repositorio: | RUO. Repositorio Institucional de la Universidad de Oviedo |
| Idioma: | inglés |
| OAI Identifier: | oai:digibuo.uniovi.es:10651/39346 |
| Acceso en línea: | http://hdl.handle.net/10651/39346 https://dx.doi.org/10.1007/s12161-016-0521-7 |
| Access Level: | acceso abierto |
| Palabra clave: | Cider apple Breeding Chemometric HS-GC SPME Volatile compounds |
| Sumario: | Gas chromatography combined with solid-phase microextraction has been used for the identification of the aromatic profiles of new cider apple genotypes, and a chemometric characterization of these new cider apple genotypes has been carried out using exploratory and modelling techniques. Three breeding targets have been explored: (1) regular bearing and scab resistance, (2) resistance to bio-aggressors and (3) high polyphenol content and late ripening. Exploratory techniques established two genotype groups: those that come from breeding towards targets 1 and 2 with low polyphenol contents and those that come from breeding towards target 3 with high polyphenol contents. Alcohols were related to the genotypes with breeding towards target 3, and compounds such as esters were related to the genotypes with breeding towards targets 1 and 2. Models computed using the soft independent modelling of class analogy (SIMCA) technique presented good sensitivity (93 %), specificity (91 %) and classification hits (96 %). However, the predictions computed by SIMCA (70 %) and the artificial neural network (ANN) (76 %) were low |
|---|