Where can we store the settings for phenotyping protocol ?

[arendd]

In the documentation of the mandatory phenotyping protocol there is only a Parameter [TDF], which can be moved to the investigation file as @proccaserra suggested in #7 and a Parameter [Observation Level]. The actual determined trait data e.g. “plant height” or ”number of leafs” is then defined in the TDF and the results stored in the different “Data files”, this is clear.

But where can we put the initial parameter settings for the phenotyping steps or rather the different images in the assay file e.g. camera configuration, recording angle, sensor setting!?

[hcwi]

For sure you can put those settings as additional parameters of the Phenotyping protocol and add necessary columns 'Parameter Value[*]' in the Assay file. Alternatively, maybe an addtional protocol e.g. 'Imaging' with these parameters can be added? If such settings are common to many high-thoughput platform, maybe there should be an additional configuration for high thoughput phenotyping assay? Worth discussing!

[arendd]

Okay, good to know. I'm not sure if we need an extra "Imaging" Protocol. I think it depends on the use cases, maybe!? I’m afraid of that this could be maybe a little bit confusing for the user if we have an “Imaging” and “Phenotyping” protocol? But this is only from our perspective, because we would probably always use both protocols, due to the fact that nearly all of our phenotyping experiments are based on imaging….

[hcwi]

You might be right about the possible confusion. Introducing "Imaging" (or other protocols related to high-throughput measurements/sensors) should be preceded by collecting info from multiple plaftorms/groups.

[proccaserra]

@arendd many moons ago :) we worked on one such example, where phenotyping was only performed via imaging techniques. There was a need to connect to the raw data so we indeed created a specific assay with an phenotypic with 'high throughput imaging' with a number of parameters (wavelength, camera angle, time of collection...). Then, appeared the need to avoid the repetition of fix values in the assay tables, and the thought of allowing for Parameter Values to have default values. This would require an alteration of the syntax, parsing rules and implementation guidelines for ISA. For HT-imaging, as it is likely to become more commonplace, is there a case for creating such specialised ISA assay configuration in addition to the one we discuss, simply to allow capturing assocations with imaging files ?