PREDICTION OF THE BREADMAKING QUALITY OF WHEAT - THE USE OF HMW GLUTENIN-A SUBUNIT-BASED QUALITY SCORING SYSTEMS

In this paper the possibilities of predicting the breadmaking quality of wheat are evaluated. This evaluation is based on the statistical analysis of two collections of Dutch-grown wheat samples. One collection, consisting of 164 wheat samples grown in 1984 is used to develop so-called prediction models. This includes models based only on HMW glutenin-A subunit compositions and an expanded model also including other parameters correlating with breadmaking quality. A second collection, consisting of 84 wheat samples grown in 1985, is used to validate the ‘1984’ models. The HMW glutenin-A subunit-based models correlate positively but poorly (r2: 0·30–0·36) with loaf volume. Extending the model with other quality related parameters (particularly SDS-sedimentation volume) leads to an improved model (r2 = 0·73). The validation using the second sample set shows, however, that the predictive value of the HMW glutenin-A subunit based model is negligible (r2 = 0·005) whereas the extended model also fails to allow a reliable prediction (r2 = 0·32). Clearly, breadmaking quality is not only governed by the quality of the HMW glutenin-A subunits, but also by other factors, one of which is correlated with the amount of glutenin proteins. This leads to the general conclusion that electrophoretic scoring systems can only be of very limited use for the prediction of breadmaking quality. © 1992, Academic Press Limited. All rights reserved.