A Streamlit app for gathering assets and mapping probe tracks in Neuroglancer to probe recordings in Open Ephys. The app then launches a data conversion capsule and creates a data asset required to use the IBL annotation app.
Credentials first need to be added to two capsules: this one, and a "monitor" capsule that handles data asset creation:
- when the
launch data converterbutton is pressed at the end of the Streamlit app, a request is made via the CodeOcean API to a "monitor" capsule - the monitor capsule starts the data cpnversion capsule (forwarding the run parameters sent from the Streamlit app)
- the monitor capsule waits for data conversion to finish (up to 3 hours) then creates a data asset from the results
- this way, the Streamlit app can be closed or used to launch conversion for another session
Credentials need to be added at the bottom of each capsule's environment page.
- AWS
- use the
AWS Assumable Role - aind-codeocean-usersecret
- use the
- CodeOcean API
- an access token is required with read and write scope on capsules and datasets (see CodeOcean docs on how to create one)
Note: These capsules should be used directly, not duplicated
- Monitor capsule: https://codeocean.allenneuraldynamics.org/capsule/5449547/tree
- Data conversion capsule: https://codeocean.allenneuraldynamics.org/capsule/8363069/tree
After adding credentials, launch the Streamlit cloud workstation (red origami icon, under Reproducible Run).
Various functions in code/extension.py are used to parse json exported from Neuroglancer, then find corresponding data assets for ecephys, spike-sorting and smartspim.
Carefully check the auto-populated tables at each stage, and remove or add lines as necessary.
It's possible to copy/paste cells from the tables, allowing you to work in Excel if you prefer.