Flavour research typically involves multifactorial-type experiments to understand the interactions between factors like raw materials, processing, chemical and sensory analyses. Developments in business, in pharmaceutical research and in genomics have led to the availability of software tools that allow large amounts of data to be analysed on one platform. Data can be text or numerical and interrogation can use combinations of both to data-mine and identify significant factors, or simply used to test hypotheses. Analysis uses the conventional statistical tools like ANOVA, PCA, PLS and clustering but, with all the data present, it is easier to interrogate the data in many different ways and visualise the outputs in 2D and 3D formats, thus making it easier to explain results to managers or clients.<br/><br/>To demonstrate the potential of these techniques, existing data on the production of volatile aroma compounds from three different cereals (oats, wheat and maize) during extrusion (different levels of moisture, time and temperature) was further analysed. The presentation will show how data can be easily imported from pdf or spreadsheet files and “data wrangled” to add other types of data (e.g. structures, molecular formulae, or physico chemical parameters like logP). After importing into one of the commercially available data visualisation platforms, examples will be given to show the power of this type of data analysis, which can represent many different parameters simultaneously, unlike the conventional 2D tables of data or the restrictions of standard spreadsheet chart-plotting tools.

Using data tools and data visualization to interpret multifactorial flavour datasets

254th National Meeting (2017)

The official archive and repository for all ACS Meetings & Expositions abstracts and select posters or presentations.