Meeting Notes

• Date: 2026-02-03
• Time: 09:00AM (PT)
• Location: Teams Meeting

Agenda

Broad update on data collection at the Allen Institute. We have started production of the final data assets.

• @jeromelecoq will present a quick recap of the experimetnal design.
• Severine Durand will share an update on where Neuropixels data collections stands.
• Shiella Caldejon will share an update on Mesoscope imaging stands.
• Maedeh Seyedolmohadesin will share an update on SLAP2 imaging progress.

We can answer questions on all of those.

Meeting Recording

Meeting Notes

Overview of Predictive Processing Data Collection Project: Jerome provided a comprehensive overview of the predictive processing project, outlining the experimental design, modalities involved, and the rationale for standardized mismatched stimuli across multiple brain scales, with contributions from Severine, Shiella, and Maedeh leading respective modality updates.

Experimental Design Recap: Jerome described the project's goal to present standardized mismatched stimuli in various contexts and measure prediction and error signals across the brain at multiple scales, including brain-wide, local, and single-cell levels, using Neuropixel probes, calcium mesoscope imaging, and SLAP2 imaging.

Cohort and Context Structure: Jerome explained the use of two mouse cohorts—one for sensory motor context and one for sequential context—trained on specific behavioral tasks, with each modality (Neuropixel, mesoscope, SLAP2) running the same experimental paradigms to ensure consistency across data sets.

Rationale and Alignment: The rationale for the experimental approach was referenced as being detailed in an archive paper, with emphasis on aligning all modalities to the same experimental design for robust cross-modality analysis.

Neuropixel Data Collection Progress and Technical Details: Severine, supported by Casey and Ryan, reported on the initial Neuropixel data collection, detailing cohort selection, session outcomes, probe targeting, optotagging results, and receptive field mapping, with active discussion involving Jordan and Stefan on technical challenges and data interpretation.

Cohort Selection and Session Outcomes: Severine described starting data collection with two mice from the motor context cohort, noting that one mouse completed all sessions and passed QC, while the other failed due to health issues; future sessions will focus on motor context before moving to sequence context.

Probe Targeting and Implantation: Details were provided on probe insertion sites, with some challenges in targeting ACA due to membrane issues, but successful recordings from MOS and MOP areas; the pilot confirmed ACA targeting via a frontal hole.

Optotagging and Spike Stability: Optotagging was performed with probe C, though responses were weaker than expected, prompting ongoing investigation; spike counts remained stable throughout experiments, allowing for reliable comparisons across sessions.

Receptive Field Mapping and Area Coverage: Severine presented receptive field data, highlighting successful recordings from V1 and LGN, and discussed with Jordan the absence of ACA receptive fields, with further technical discussion on stimulus optimization and mapping challenges.

Data Release and Documentation Plans: Alexander Maier inquired about data release timelines and documentation, with Jerome and Carter indicating that data uploads are expected next week, and plans to improve documentation and provide experiment summaries for user guidance.

Mesoscope Imaging Data Collection Update: Shiella, with support from Jimin, presented the status of mesoscope imaging, describing the experimental setup, cohort details, imaging targets, and technical specifics, while addressing questions from Jordan and Jerome regarding imaging planes and acquisition methods.

Experimental Setup and Cohort Details: Shiella reported using two-photon mesoscope systems to target V1 and LM across layers 1-5, currently working with one animal from the motor cohort, and reviewing QC to ensure data quality before proceeding.

Imaging Planes and Acquisition: In response to Jordan's questions, Shiella explained that paired planes are imaged simultaneously during each session, with temporal multiplexing and remote focus enabling acquisition of eight planes at a volume rate of 8-10 Hz.

Team Roles and Collaboration: Shiella introduced herself as the RA Senior Supervisor for the Optical Physiology team, with Jimin also contributing to data collection and project support.

SLAP2 Imaging Data Collection and Technical Discussion: Maedeh provided an update on SLAP2 imaging, detailing the custom microscope's capabilities, experimental progress, and future plans, while engaging with Alexander Maier, Lucas, and Stefan on technical aspects such as input-output measurements, crosstalk, and indicator properties.

SLAP2 Microscope Capabilities: Maedeh described the SLAP2 microscope as a custom-built, random-access two-photon system capable of recording hundreds of synaptic inputs and somatic outputs at high spatial and temporal resolution, targeting perisomatic and apical dendrites in layer 2-3 pyramidal cells.

Experimental Progress and Indicators: Three animals expressing glutamate and calcium indicators have been recorded, with a fourth planned; future experiments will use ASAP family voltage indicators to measure subthreshold membrane potentials, providing complementary views of neuronal computation.

Input-Output Measurement Challenges: In response to Alexander Maier, Maedeh confirmed that calcium signals from the soma are recorded alongside glutamate inputs, though calcium indicators have shorter survival times, allowing input-output measurements for 20-30 minutes per session.

Crosstalk and Source Extraction: Lucas raised concerns about potential crosstalk between synapses during event-based analysis; Maedeh explained that the preprocessing pipeline uses validated source extraction methods to minimize crosstalk, though minor overlap may still occur.

Indicator Affinity and Measurement Impact: Stefan asked about the affinity of the snfr indicator and its potential to outcompete native channels; Maedeh indicated that more information is available in the relevant publication, and Jerome suggested consulting Kaspar's recent paper for further details.

Cross-Modality Data Alignment and Analysis Planning: Jerome and Alexander Maier led a discussion on ensuring alignment across Neuropixel, mesoscope, and SLAP2 modalities, confirming that all use the same experimental design with minor variations, and proposed steps for collaborative analysis, documentation, and community engagement.

Experimental Consistency Across Modalities: Jerome confirmed that all three modalities—Neuropixel, mesoscope, and SLAP2—are running the same experiments, with only minor differences such as SLAP2's initial use of single zebra presentations.

Analysis Pipeline and Community Collaboration: Alexander Maier emphasized the need for analytical code and clear documentation to facilitate data use by the broader community, suggesting the creation of forum threads to outline completed and planned analyses, and steps for working with NWB data.

Leadership and Organizational Structure: Jerome proposed starting a discussion thread to organize analysis efforts, inviting group members to contribute ideas and help structure collaborative work as the project transitions from data collection to analysis.

Technical Questions and Data Interpretation: Participants including Jordan, Stefan, Lucas, and Alexander Maier raised and discussed technical questions regarding receptive field mapping, stimulus optimization, data crosstalk, and indicator properties, with responses provided by Severine, Maedeh, and Jerome.

Receptive Field Mapping in ACA: Jordan inquired about ACA receptive fields, with Severine confirming none were observed and discussing possible reasons, including stimulus properties and anatomical targeting.

Stimulus Optimization for Mapping: Stefan suggested that mapping stimuli may not be optimized for large receptive fields, prompting discussion on the use of broader parameter sets and the need for specialized analysis.

Crosstalk in SLAP2 Data: Lucas asked about potential signal bleed between synapses in SLAP2 data, with Maedeh explaining that validated source extraction methods minimize crosstalk, though minor overlap may still occur.

Indicator Affinity and Measurement Impact: Stefan raised concerns about the snfr indicator's affinity and its impact on neuronal output, with Maedeh and Jerome recommending further review of published documentation for detailed properties.