Discussion: meetings/2026-09-01

[jeromelecoq]

Post your material here

[Sarruedi]

We have been working on the electrophysiology quality-control figure from two complementary angles:

@yhcheah804 has focused on developing an interactive interface for exploring the electrophysiology data and QC metrics.
@mariannasidiropoulou has been working on spike-waveform classification, both within individual brain areas and across areas using wavemap.

The resulting HTML dashboard for the QC metrics is unfortunately too large to upload here, but we would particularly like to discuss the interface and visualization strategy, CCF/anatomical alignment with probe depth, QC criteria to use as we refine the analysis and decide on the next steps to finalize a draft the figure for the Ecphys panels.

[TimStamm]

I haven't added any new analyses yet. I've mostly been working on a data-analysis app to bring the project's existing QC and analysis workflows together in one place.

The idea is to integrate the collaboration's existing methods rather than replace them. I'm currently incorporating mesoscope QC from the OpenScope ophys-QC repos (@Alishamsnia, @Dedalus9 , @davisgrubin) , along with RF, tuning, oddball, and other analyses from the collaboration repos, and have started on the Neuropixels QC workflows. I'll attach a few slides showing the current app and workflow.

It would also be helpful to have a pinned discussion/resource listing the project's processing and analysis repos, notebooks, and Code Ocean capsules, with a short description of what each is used for.

A few related questions (@jeromelecoq):

1. Earlier analyses seemed to work primarily from the S3 assets, while the data-release work now seems to use the complete DANDI NWBs as the analysis-ready source. Is DANDI generally the preferred source going forward when a complete NWB is available?
2. Working directly from the mesoscope S3 assets required some DAQ/Bonsai/BonVision timing alignment. Am I correct that this synchronization has already been incorporated into the complete DANDI NWBs? Is there a preferred repo/capsule containing the synchronization implementation?
3. The S3 sessions also have several _processed products. I've found some Allen repos/capsules that appear to generate these, but I'm not sure I have the complete processing lineage. Could those processing repos/capsules also be included in the resource list?

Longer term, I'd like the app to cover as much of the reproducible path as practical:

raw/source data → processing/synchronization → QC → scientific analyses → figures/results

Any suggestions for features, interface flow, etc. are welcome.

Note: I'll try to attend as much of the meeting as I can, although I still don't have a mic on my work computer yet.

OpenScope_Integrated_Analysis.pptx

[RobertoDF]

I checked whether in the neuropixel dataset sensory motor oddballs happen during running or not upon @jeromelecoq suggestion, and indeed this seems to be the reason why there arent strong responses. The only 2 mice where there are clear responses to the motor halt oddball happen to be the ones that actually run during the oddball. I guess we should show something abt this to explain absence of responses?
mesoscope and SLAP seem to have a bit more running going on
https://robertodf.github.io/openscope_p3_data_release_paper/dev/running_summary_all_mice.html

[davisgrubin]

Hi there, just wanted to provide a quick update on progress on Figure 7 Mesoscope SNR by ROI QC. I watched the last meeting with Ido's presentation of the SLAP2 panel and have been working on a comparable version for the Mesoscope data but have not finished that yet, should have figures to present by next week's meeting though.

[jeromelecoq]

Sounds good, I will add you to the schedule next week then.