OpenAg basil experiment data from our food server


This is one of our bio team’s recent experiments and all the data that was collected in one static repo for easy reference.

pinned globally #2

@rbaynes I’m excited to see this data! Thanks for posting.

Was MS-GC data collected for this experiment? I just read the paper which OpenAg published and from what I can tell that appears to be referencing a previous experiment in the “V1 Food Server” prior to Bates. That being said, I was very intrigued by the metric which was established of “Chemscore” and the identification of 24-hour photoperiod being ideal for maximizing that metric.

  1. What was the photoperiod used for this dataset, that information appears to be missing and I would argue given the previous findings that it is critical.
  2. The fields for tissue ppm and % are not referenced in the meta file, and the data appears to be the same for all plants in a bay. Due to the variance in DLI between plants in a given bay, it would be helpful to know which plant these tissue samples were taken from.
unpinned #4

@Webb.Peter @hildreth
Hi! This is Rebekah, one of the plant scientists on OpenAg team. First of all, thank you for reading the paper and looking into our data! It is exciting to share our work with everyone.

As far as your questions,

  1. The photoperiod is described in the OpenAg Data Road Map from 6:00 am to 10:00 pm, so we used a 16 hr photoperiod.

  2. And thanks for highlighting the tissue and % ppm overlook on the meta file, I’ll include it and update the file soon. In the meantime, for tissue nutrients, we took a subsample of 6 plants per bay and combined them to do an average of the nutrients. Because this experiment was designed around the larger effect of rack design, rather than as granular as each plant, we did not do individual nutrient work ups.

1 Like

@rcarlson I appreciate the response, I know you all are very busy.

  1. I see it now, my apologies. It may just be that it wasn’t included in the export but it would be nice to have some indication of this perhaps in the data itself. Perhaps if the data were structured to provide the umol measurements for every time the lights are on and 0 for when they’re off it would make this more friendly to direct export or integration long-term.
  2. No worries! This makes sense, I figured it was an average but wanted to check and try to help out. Just out of curiosity, what was the variable in question with regards to system design?

Thanks for the perspective on the data, Peter! To your second point, we were testing the reservoir design in this experiment… determining if there was any difference in plant response using a 20 gallon reservoir system versus a 40 gallon one.