Discussion: meetings/2026-08-25

[jeromelecoq]

Post your slides here

[ido4848]

SLAP2 glutamate column of Figure 7 (panels C, G, K) is in #171, discussion in #172.
Event detection at 3 SD, per-synapse SNR classes, 20 sessions / 8 mice.

C: example session, ROIs coloured by class

G: event amplitudes and SNR across the archive

K: event rate by stimulus context, per class

[RobertoDF]

Here I focus only on Sensorimotor mismatch responses in neuropixel sessions:

• motor halt seem to elicit very small responses (strong only in a couple animal, Keller uses 1 sec long mismatch, here 343 ms), omission causes more clear responses and orientation change even stronger.
• motor omission responses are statistically different in the motor and sequence cohort
• layer 2/3 SST responses to omission and orientation change are statistically different in the motor and sequence cohort

here 2 interactive visualizers:
PSTHs
https://robertodf.github.io/openscope_p3_data_release_paper/dev/SDF_inspector_population.html
(these are simple PSTH, not z-scored, calculated with spike density function similarly to https://www.nature.com/articles/s41586-020-03171-x)

summary split by cohort
https://robertodf.github.io/openscope_p3_data_release_paper/dev/SDF_inspector_summary.html

[Sarruedi]

Hi Roberto,

Thank you for posting your interactive code. This looks very promising!

One quick suggestion, it would be great to visualize your psth plots using a wider x-axis. The current plots are quite compressed along the x-axis (at least on my display). Would it be possible to position the legend elsewhere so you have more space available for the actual plot?

[RobertoDF]

Does it look different than this in your display?

[Sarruedi]

Yeah, mine is much more squished. maybe it's because I looked at the graphs using a cell phone:

[RobertoDF]

ah right, its definitely because of the phone.

[mariannasidiropoulou]

Sharing our Ecephys slides (4 for the 3 of us) for tomorrow’s meeting, put together with Sarah (@Sarruedi) and Yuheng (@yhcheah804) focusing on Figures 7E & 7I of the data-release manuscript. Looking forward to discussing these with everyone tomorrow! 😊

[davidgwyrick]

Hello team! I also made an interactive PSTH viewer like Roberto, but it has the option to display all of the different stimulus contexts with various unit inclusion criteria that one might be interested in. I would post the interactive version, but I think my code is a branch on the main repository, not a fork, so the github page publish doesn't work. The static version of the figure needs some work, but I thought the PSTH viewer was a nice way to look at various mismatch responses without quantifying any sort of "oddball index" or anything yet. Although, that is easy to do for a static figure if that falls within the purview of the data release paper.

Neuropixels mismatch-response explorer_sensorimotor.pdf

[davidgwyrick]

[antonio-velazquezr]

Fig 1. LFP spectral analysis in VISp2/3: example raw trace + spectrogram showing 25-35Hz burst around stimulus, power spectra by condition (orientation stimuli > standard > omission, with significant band highlighted), and quantified 25-35Hz power (Hilbert AUC) — both oriented gratings show significantly elevated power vs standard, omission significantly reduced, halt no different.

Fig 2. SST+ tagged units vs non-tagged: light-tagging validation, example LFP trace with 25-35Hz filtered spikes, population firing rates (standard vs oddball), example phase-locking polar plots, and PPC quantification — tagged and non tagged units show significantly stronger 25-35Hz phase-locking under oddball conditons (***, Wilcoxon).

TL;DR: A 25-35Hz rhythm is stimulus-driven (strongest for oriented gratings, suppressed on omission) and SST+ increased their firing rate and also their coupling of themselves and non tagged units

[jeromelecoq]

Sorry @antonio-velazquezr we did not get to see your analysis. We should make sure we start with yours next week.