Fractional mixture design: a novel design and methodology for analyzing and optimizing mixtures of several ingredients.

Abstract

Mixture designs are employed to systematically change the composition of mixtures and investigate how those changes impact their properties. However, all mixture designs currently available are impractical for analyzing mixtures with relatively large numbers of ingredients. In response, this article presents a novel solution that builds on the construction of a new experimental design called "fractional mixture design". The design involves screening the ingredients in mixtures and enables the subsequent construction of a classical mixture design for optimizing mixtures. The design and its accompanying methodology were developed to analyze native strains found in successful spontaneous fermentations with the goal of constructing a mixed starter culture to transition from spontaneous to directed fermentation in the production of agave distillates. The results showed that a starter culture composed of the native strains Kluyveromyces marxianus, Clavispora lusitaniae, and Kluyveromyces marxianus var. drosophilarum, in respective proportions of 35%, 32%, and 33%, enabled the production of a fermented product with 2.1% alcohol and a broad profile of aromatic compounds. Hence, the results show, for the first time, a tool that addresses the technical challenge that allows studying a relatively large number of ingredients in mixtures and a two-stage sequential methodology to construct optimal mixtures.