Meeting Notes

• Date: 2025-05-06
• Time: 9:00 (PST)
• Location: Teams Meeting
• Facilitator: @jeromelecoq @IBussi @hbelski

---

Meeting Recording

Agenda

1. Walk through project website
2. Brief recap of the project goals
3. Discussion on collaboration guidelines Goal: Agree on guidelines
4. Discussion on how we track progress Goal: Share how we will track progress at the Allen Institute. On board more teams.
5. Discussion on Stimulus design and deployment Goal: Align on timeline and sub-groups involved.
6. Discussion on Analysis plans and tracking Goal: Align on list of analysis projects and facilitate sub-groups formation.

---

Notes

Introductions and Project Overview

• First public kick-off for the Predictive Processing project.
• Round of introductions from the Allen Institute team.
• Walkthrough of the project website

Experimental Design

• Many community members are collecting data independently in their own labs.
• Kaspar asked whether the experimental design is finalized.
  • Jerome clarified: protocols defined and carried out by OpenScope are already defined. However, satellite experiments can still evolve.
• Consistency across labs will be a challenge, but using shared stimuli helps ensure alignment.

Collaboration Policy

• Jerome walked attendees through the collaboration policy.
• Kaspar raised a question about authorship order in data set that were produced by independent labs.
  • Publications will be led by contributing groups; if disputes arise, the Steering Committee will help resolve them.
• Julien asked how to contribute experimental data.
  • Jerome encouraged contributors to reach out. If data has already been collected, it can be added to the website with an explanation of how it complements the existing dataset.
• Ryszard mentioned their team is contributing to the analysis plan.
  • Jerome noted that current analysis tickets are mostly placeholders, but people can begin contributing directly on the GitHub board.
• Karim asked about incorporating stimulus variations and how to determine whether such data falls under the umbrella of the project.
  • Jerome: If the project is listed on the website, labs should strive to align with the experimental design or clearly describe any deviations.
  • Iterative updates and variations are welcome as long as they're documented and justified.

Project Tracking

• The team at the Allen is using a Kanban board to track tasks, which includes many placeholders to be refined over time.
• External labs are encouraged to use the same system to promote transparency and facilitate coordination.

Stimulus Design & Deployment

• Stimulus development resources are available in the “Project Resources” section of the website.
• Scripts were developed in Bonsai, and Josh Siegle wrote a guide comparing Bonsai and Python.
• Four required stimuli, three are currently under development.
• The standard mismatch oddball script is nearly complete, with some data already collected and available in NWB format on the website.
  • Jerome collaborated with Kaspar and Maedeh to adapt the script for SLAP2 systems.
• Plan to schedule a biweekly stimulus design meeting for ongoing coordination.
• Andre commented they are working on adapting the protocol to primates. Requested feedback on how to tailor the proposed paradigm. One potential approach involves using joysticks.
  • Jerome: Several joystick-controlled stimulus possibilities; ideas still developing. Encourage people to use the discussion windows to start a thread on the topics that need more discussion.
• Alex and Karim: Discussed gain variations and speed control in stimuli.
  • Jerome confirmed gain is easily modifiable in the script.
• Any changes agreed upon in discussion should then be added to the relevant pages on the website.
• Jerome added access control to the main branch of the repo. Interested contributors can request editing permissions.

Analysis Plan and Coordination

• The project focuses on three main hypotheses, each with associated analyses.
• Coordination needed as data comes in to avoid duplication of efforts.
• SLAP2 dataset is available on DANDI, ready for analysis.
  • Preliminary data shows dendritic responses to oddball stimuli.
  • Results to be presented in upcoming meetings.
• Alex suggested creating a discussion page for analysis to facilitate feedback and idea sharing.
• Kaspar suggested to organize experiments on the website into groups rather than individual listings.
• Jerome encourage people to sue mkdocs.yml to add entries to the website and reorganize the experiments.

Next Steps and Action Items

• Next meeting 5/13: Stimulus design discussion.
• Following meeting 5/20: Deep dive into analysis plan, including SLAP2 data.