This work fits into the context of cassava processing. Production and consumption of this product is steadily increasing worldwide and especially in tropical regions where, after harvest, cassava starch is extracted according to an empirical process: natural fermentation and sun-drying, which gives to this product an interesting breadmaking capacity despite of the absence of gluten. The objective of this work is to try to explain the breadmaking ability from different parameters (physicochemical and spectroscopic data) using a statistical regression method while selecting variables of different types: individual and intervals. In chemometrics, the choice of explanatory variables is a problem often discussed, but, when it comes to select intervals, methodologies are rarer and more complex (Höskuldsson 2001). Among the specific methods developed (Norgaard et al 2004), Genetic Algorithms (GA) were chosen and combined with the PLS method to select intervals (Leardi 2000, Leardi and Norgaard 2004). In this case, the explanatory variables are organized in a multitable in which intervals and individual variables are selected in order to predict one variable of interest: the breadmaking capacity. To this end, we will use and adapt a GA developed in a context of discrimination (usual LDA), jointly with the PLS1 method (Reynes et al 2006), this method is called AGvPLSm.