Philip Durbin 🚀 (Sep 03 2025 at 18:10): Philip Durbin 🚀 (Sep 03 2025 at 18:10):Hi @Svetlana Lebedeva I was just telling @Germán Saracca about how you mentioned some pain points you're having entering information about biological samples. I know the conversation was recorded (around 14:00) and there are notes but if you don't mind writing a bit here around the problem, we'd be happy to understand your feedback better!
Philip Durbin 🚀 (Sep 03 2025 at 18:20):Here's the transcript: transcript.txt
Philip Durbin 🚀 (Sep 08 2025 at 16:01):@Svetlana Lebedeva you mentioned FAIRDOM-SEEK, right? It looks like their website is https://seek4science.org and https://github.com/seek4science/seek is their GitHub. They have this feature already? The biologist fills in an Excel sheet with multiple samples and uploads it?
Philip Durbin 🚀 (Sep 08 2025 at 16:02):"Samples – FAIRDOM-SEEK has a flexible samples framework that allows users to generate templates for sample metadata. Samples can be data, or laboratory samples. The framework integrates with Rightfield sheets, so that semantically-enhanced title sheets can be included with the data. The default Rightfield sheet uses the JERM and allows automatic generation and linking of the data files to an assay. Sample attributes can be limited to controlled vocabularies backed by an ontology." -- https://seek4science.org/features.html
Svetlana Lebedeva (Sep 10 2025 at 12:26):hey! Sorry, I was not configured to receive notification for mentions.. Yes, you described it precisely - ideally researcher fills out an excel sheet, allowing to copy-paste, autofill, etc., which then should get ingested. Usually we will have ~10-20 samples associated with a dataset, sometimes more. Right now we handle each sample as a compound field, allowing multiples of the samples, and building an external tool which will provide excel-like functionality - because obviously, nobody will type 20 samples x 6-8 fields by hand...
Screenshot from 2025-09-10 14-20-48.png
Philip Durbin 🚀 (Sep 10 2025 at 13:16):Are you saying that the manual way for the researcher to enter the samples would be to click "add" 6-8 times and then fill in the details for each sample?
Svetlana Lebedeva (Sep 10 2025 at 13:47):Yes, I couldn't come up with any alternative of uniting multiple samples under a single dataset, but of course happy to accept ideas :-)
Philip Durbin 🚀 (Sep 10 2025 at 13:49):So your plan is to have researchers run a script to populate those 6-8 samples?
Philip Durbin 🚀 (Sep 10 2025 at 13:51):Oh, I have the numbers wrong. ~20 samples per dataset. 6-8 fields per sample.
Philip Durbin 🚀 (Sep 10 2025 at 13:51):Actually, not a script, right? An external tool? Something in the browser?
Svetlana Lebedeva (Sep 11 2025 at 15:38):Right, something written in python which will give them excel-like graphical user experience and send the data to dataverse via API in the backend. The Sample fields should stay hidden in the actual dataverse UI at the time of dataset creation so the user is not confused.
Philip Durbin 🚀 (Sep 11 2025 at 15:40):A Python webapp?
Oliver Bertuch (Sep 12 2025 at 05:23):@Svetlana Lebedeva just in case you haven't heard about it: at FZJ an institute developed https://github.com/sciapp/sampledb. It has a push button solution to push into Dataverse via API. Maybe good for inspiration or even reuse?
Svetlana Lebedeva (Sep 12 2025 at 09:30):Thanks for the tip! Will look into this :)
Oliver Bertuch (Sep 12 2025 at 10:38):Feel free to contact me if necessary. https://go.fzj.de/obertuch
Philip Durbin 🚀 (Sep 12 2025 at 10:46):Mmm, export to Dataverse: https://scientific-it-systems.iffgit.fz-juelich.de/SampleDB/administrator_guide/dataverse_export.html :yum:
Last updated: Nov 01 2025 at 14:11 UTC