Gluten aggregation properties extrapolation for a comprehensive evaluation of wholegrain and refined flours

A large amount of rheological device is used to investigate the physical properties of doughs at research and companies level. Among them, GlutoPeak (Brabender, Duisburg, Germany) has been recently proposed for the evaluation of wheat flour quality by measuring the aggregation behaviour of gluten. The proposed device goes beyond other methods limits, being time saving (5 min) and requiring small amount of sample (< 10g). Several indices can be automatically extrapolated from the GlutoPeak profile, which have demonstrated to be significantly correlated with the conventional parameters used for wheat characterization and dough-quality prediction. However, the hidden information within the entire profile obtained by the GlutoPeak analysis has never been evaluated itself. Moreover, GlutoPeak has never been proposed for the analysis of wholegrain flour to determine process efficiency (e.g. dough yield, leavening conditions, and so on) or end-product quality (e.g. specific volume, textural attributes). The proposed work aims at demonstrating that it is possible to extrapolate relevant information from both wholegrain and refined flours by evaluating the entire Glutopeak profile by chemometric techniques. To the aim 44 GlutoPeak profiles were collected from 22 wholegrain wheat samples and the related refined flours for being analysed by Principal Component Analysis (PCA). PCA scores well discriminate wholegrain samples with high Farinographic Stability (FS) (≥ 18.5 min) from low FS samples (<18.5 min) as well as the related refined flours. The evaluation of PCA loadings assessed that sample grouping is ascribable to the combination of variables 73s and 148s and 37s and 70s for wholegrain and refined flour, respectively. Furthermore, k-nearest neighbour (k-NN), a non-parametric method used for classification, classified both wholegrain and refined flour from the Glutopeak profile according to their FS. In particular, the cross-validated average prediction rate was 81.82 % for both the calculated models (wholegrain and refined flours). The results obtained from the screening of a limited number (22) of wholegrain wheat and the related refined flours are encouraging in showing how GlutoPeak profile could be used as a fast approach for predicting wholegrain FS and thus flour performance.