Oregon State University, Corvallis, United States
The objectives of this study were to examine new approaches to identify gluten proteins with enhanced impacts on noodle texture, and to develop prediction models (PMs) for noodle texture. Wheat flours with protein contents from 9.1–13.9 % that were harvested in Oregon in 2003 were used for the study. Samples were typed by HMWGS compositions, SE-HPLC chromatography of proteins, and Mixograph and RVA pasting characteristics. Continuum regressions were applied to the chromatograms to give primary indications of relationships between gluten proteins of different Mr and mixograph and cooked noodle attributes. As an extension of these observations, principle component scores (PCS) were calculated from absorbance and % absorbance values from SE-HPLC, and from RVA pasting data. The PCS and their cross products, ratios, and squares were used to calculate the PMs, both with and without the RVA data. PMs for mixing and noodle characteristics showed better performance when calculated from PCS values of both SE-HPLC and RVA data than from SE-HPLC data only. The R2 values of PMs for mixograph absorption, peak time, and tolerance were 0.827, 0.813, and 0.851, respectively. The PMs for noodle hardness, cohesiveness, chewiness, and resilience had R2 values of 0.928, 0.928, 0.896, and 0.855, respectively. These results indicate the potential of multivariate regression methods, using SE-HPLC and RVA data, for the development of robust PMs for dough mixing and noodle characteristics. The PMs could be applied in circumstances where insufficient material was available for noodle or dough-testing.