Fig. 3

Overview of growth assay workflow including laboratory execution and data analysis. A technician at the lab bench is guided through the steps of collecting images of containers of duckweed according to the plan launched for the experiment (A). The images are uploaded via an on-screen prompt by the technician as part of the Aquarium job. The technician takes an image of the plate and uploads it to Aquarium. A Jupyter notebook guides the user through the process of identifying the plate borders, duckweed in the plate, and quantifying percentage coverage (B). In the strongly typed Aquarium LIMS, all objects have a unique ID as well as a number of possible properties with unique names (C). This system is the basis for the design of analysis scripts using the Trident API. For example, our data analysis script relies on user input of a specific plan ID to identify an experiment, then image data are extracted by performing a scripted search for Items of type ‘Container of Duckweed’ and pulling all data associations from each item. We also include an operation of type ‘Define plate treatment’ in each experiment plan to store important metadata, extracted in a similar way using the unique name or ID of the operation type to find it within the operations associated with the plan of interest