Details of talk

TitleA talk about beer, wine, and pizza
PresenterAmelia Gontar (Flinders University)
Author(s)Amelia Gontar
SessionContributed Talks
Time11:00:00 2017-09-26
Abstract


The yeast \textit{Saccaromyces cerevisiae}, often called bakers' yeast, is used
in the production of bread, wine, ale, and -- of course -- pizza dough.
\textit{S. cerevisiae} can grow via the budding of yeast cells, or by forming
chains of unseparated cells called pseudohyphae. Pseudohyphal growth is thought
to be triggered by nutrient deprivation, but is not yet completely understood.
Hence, quantifying the often complex shape patterns exhibited by pseudohyphae is
of interest. In this project, pseudohyphal growth is analysed using
two-dimensional top-down binary images. The colony morphology is characterised
using clustered shape primitives -- shape descriptors learned automatically from
the data. Features based on clustered shape primitives are used to build a
classifier to distinguish between yeast colonies based on strain, nutrient
concentration, and by both strain and nutrient concentration simultaneously. The
classifier uses only a small set of features that are learned automatically from
the data, and does not require a list of predefined features or \textit{a
priori} knowledge about the data sets. The resulting classifier predicts the
strain and nutrient concentration of the yeast colonies with an accuracy
of~$0.969$ ($\text{SD}=0.041$), demonstrating that yeast colonies exhibiting
pseudohyphal growth can be characterised successfully based on their shape
patterns.

Back To Timetable