The Mathematical Biology of Diatoms

6Diatom Growth: How to Improve the Analysis of Noisy Data

Olga Kourtchenko¹*, Kai T. Lohbeck², Björn Andersson¹ and Tuomas Rajala³

¹ Department of Marine Sciences, University of Gothenburg, Gothenburg, Sweden

² Limnological Institute, University of Konstanz, Konstanz, Germany

³ Applied Statistical Methods, Natural Resources Institute Finland, Helsinki, Finland

Abstract

Measuring cell culture growth is a standard phenotyping technique in microbial research. It allows the determination of key parameters, such as exponential growth rate, acclimation time, and yield, which have been used extensively to study microorganisms’ gene function, physiology, and evolution. Diatoms, being unicellular eukaryotes, are commonly phenotyped in the same fashion as other microorganisms. However, the resulting growth data is distinct in many ways, making it challenging to apply standard parameter estimation methods used, for example, for heterotrophic bacteria. Diatom growth data analysis often translates to the manual, ad hoc, data fitting, typically involving linear regression for exponential growth rate. The rate value is subsequently regressed against the control factors in an experiment. The quality of rate estimate thus determines the meaningfulness of the following regression analysis, making the selection of the appropriate (most robust, accurate, and objective) method for rate calculation of utmost importance.

We explored a suite of mathematical methods for specific growth rate estimation ...