Meeting Notes
- Date: 2025-06-17
- Time: 09:00 (PT)
- Location: Teams Meeting
- Presentations: @Dedalus9 @jeromelecoq @Sarruedi
Agenda
All content should be shared one day prior to the meeting on the linked forum thread (click at the bottom of this page to upload).
- SLAP2 analysis data snapshot 1: Nicholas
- SLAP2 analysis data snapshot 2: Sarah
- Discussion on stimulus design iteration for standard oddball : Jerome and everyone
Meeting Recording
Meeting Notes
Meeting Agenda: Jerome outlined the meeting agenda, which includes presentations by Nicholas, Sarah, and Jerome on various analyses and experimental designs.
Agenda Overview: Jerome outlined the meeting agenda, which includes presentations by Nicholas on population level analysis, Sarah on individual ROI analysis, and Jerome on experimental design variants.
Presentation Details: Jerome mentioned that Nicholas would present analysis on slap two data, Sarah would present her findings, and Jerome would discuss different variants of the experimental design conducted.
Discussion Format: Jerome explained the format of sharing presentations prior to the meeting to facilitate in-depth discussions during the meeting.
Population Level Analysis: Nicholas presented his population level analysis on slap two data, discussing preprocessing steps, tuning preferences, and the impact of photo bleaching on amplitude.
Preprocessing Steps: Nicholas described the preprocessing steps, including normalization by baseline F0 and selection of ROIs based on orientation tuning preference changes after the oddball block presentation.
Photo Bleaching Impact: Nicholas discussed the impact of photo bleaching on amplitude and how he addressed it by focusing on the shape of the ROI tuning vector rather than amplitude.
Tuning Preferences: Nicholas explained how he calculated the tuning spread per ROI and performed min-max scaling to address photo bleaching, focusing on the shape of the ROI tuning vector.
Orientation Tuning: Nicholas explained how he calculated the tuning spread per ROI and performed min-max scaling to address photo bleaching, focusing on the shape of the ROI tuning vector.
Tuning Spread Calculation: Nicholas calculated the tuning spread per ROI by determining the maximum and minimum values for each peristimulus window and retaining ROIs in the top 25% of the tuning spread distribution.
Min-Max Scaling: Nicholas performed ROI-wise min-max scaling to rescale every orientation tuning vector between zero and one per ROI to address photo bleaching.
Orientation Tuning Vector: Nicholas focused on the shape of the ROI tuning vector to mitigate the impact of photo bleaching on amplitude measurements.
Radar Plots: Nicholas shared radar plots to visualize changes in orientation tuning after the presentation of the oddball block, noting patterns of multiple stimuli tuning and dominant tuning changes.
Radar Plot Visualization: Nicholas shared radar plots to visualize changes in orientation tuning after the presentation of the oddball block, noting patterns of multiple stimuli tuning and dominant tuning changes.
Multiple Stimuli Tuning: Nicholas observed that certain ROIs exhibited tuning to multiple stimuli and changed their tuning after the oddball block presentation.
Dominant Tuning Changes: Nicholas noted that some ROIs had one dominant tuning that changed significantly after the oddball block presentation.
PCA Analysis: Nicholas used PCA to analyze changes in orientation tuning vectors, aiming to identify dominant covariance modes and patterns of change across multiple orientations.
PCA Methodology: Nicholas used PCA to compute the cross covariance matrix between pre and post oddball block orientation tuning vectors to identify dominant covariance modes.
Covariance Modes: Nicholas aimed to identify whether changes in orientation tuning vectors exhibited dominant covariance modes focused on single or multiple orientations.
Pattern Analysis: Nicholas analyzed patterns of change across multiple orientations using PCA, noting that changes were not focused on a single orientation but rather across multiple orientations.
Plotting Issues: Lucas and Alexander raised concerns about plotting issues related to orientation angles and the use of direction versus orientation in the analysis.
Orientation Angles: Lucas and Alexander raised concerns about the accuracy of orientation angles in the radar plots, suggesting potential issues with the conversion of direction to orientation angles.
Direction vs Orientation: Lucas and Alexander discussed the difference between direction and orientation angles, highlighting the need to ensure correct computation and plotting in the analysis.
Control Trials: Nicholas emphasized the need for control trials to enable a more robust population level analysis and address potential biases in the PCA process.
Need for Controls: Nicholas emphasized the importance of control trials to enable a more robust population level analysis and address potential biases in the PCA process.
Bias Concerns: Nicholas discussed concerns about potential biases in the PCA process and the need for control trials to mitigate these biases.
Gain Modulation: Sarah presented her analysis on individual ROI level, highlighting a marked gain modulation of synaptic strength in Block 2 compared to Block 1.
Synaptic Strength: Sarah highlighted a marked gain modulation of synaptic strength in Block 2 compared to Block 1, noting that the response amplitude in Block 2 was consistently smaller.
Response Amplitude: Sarah observed that the response amplitude of Delta F/F was almost always smaller in Block 2 compared to Block 1, indicating a gain modulation effect.
Normalization Method: Sarah explained her normalization method, peak normalizing responses to the maximum response across both blocks to illustrate gain changes.
Orientation Selectivity: Sarah discussed the orientation selectivity index for each ROI, noting that while most ROIs remain stable, some exhibit dramatic changes in selectivity.
Selectivity Index: Sarah discussed the orientation selectivity index for each ROI, noting that while most ROIs remain stable, some exhibit dramatic changes in selectivity.
Stable ROIs: Sarah observed that the majority of ROIs remained stable in their orientation selectivity index across blocks.
Dramatic Changes: Sarah noted that some ROIs exhibited dramatic changes in their orientation selectivity index, either losing or gaining selectivity.
Preferred Orientation: Sarah visualized the preferred orientation for each ROI, showing that while many ROIs do not change their preference, some exhibit significant shifts.
Preferred Orientation: Sarah visualized the preferred orientation for each ROI, showing that while many ROIs do not change their preference, some exhibit significant shifts.
Stable Preferences: Sarah noted that many ROIs did not change their preferred orientation across blocks.
Significant Shifts: Sarah observed that some ROIs exhibited significant shifts in their preferred orientation, indicating changes in tuning.
Response Strength: Sarah highlighted the consistent change in response strength across sessions, with many ROIs showing a scaled-down response amplitude in Block 2.
Consistent Change: Sarah highlighted the consistent change in response strength across sessions, with many ROIs showing a scaled-down response amplitude in Block 2.
Scaled-Down Amplitude: Sarah observed that many ROIs showed a scaled-down response amplitude in Block 2 compared to Block 1.
Session Consistency: Sarah noted that the change in response strength was consistent across different sessions, indicating a robust effect.
Receptive Field Mapping: Participants discussed the importance of receptive field mapping as a sanity check and potential internal control for plasticity analysis.
Sanity Check: Participants discussed the importance of receptive field mapping as a sanity check to ensure the accuracy of synaptic input localization.
Plasticity Control: Participants considered receptive field mapping as a potential internal control for analyzing plasticity effects in synaptic inputs.
Experimental Design Variants: Jerome presented several experimental design variants to control for learning effects and contextual modulation, seeking feedback on the most important controls.
Statistical Power Analysis: Lucas and Nicholas agreed to collaborate on statistical power analysis to determine the number of trials needed for reliable measurements in different blocks.