Fig. 1: From experimental design to response patterns with Kinbiont.
From: Translating microbial kinetics into quantitative responses and testable hypotheses using Kinbiont
Microbial kinetics experiments generate time-resolved datasets under controlled conditions. These datasets serve as input to Kinbiont. The pipeline consists of three modules: (1) optional data preprocessing, including background subtraction, smoothing, or optical-density corrections tailored to microplate assays; (2) model-based parameter inference to characterize growth curves quantitatively (see Fig. 2 for examples of supported models and analyses); and (3) glass-box machine learning methods to perform feature selection and establish interpretable relationships between inferred growth parameters and experimental variables. The output from this module is either an empirical mathematical law (via symbolic regression) or a graphical decision model (e.g., a decision tree), describing how experimental conditions influence microbial growth parameters. All data shown in this figure are synthetic and intended solely for illustrative purposes.
