Agenda

From Sarah Ruediger's group:

  • Manav Ponnekanti : Receptive Field analysis of SLAP2 data
  • Ogulsuray Hydyrova : Experience dependent plasticity of Orientation tuning

Meeting Recording

Meeting Notes

Analysis of Slap2 Synaptic Activity Data: Manav presented a comprehensive analysis of pilot Slap2 synaptic activity data collected using Kaspar Podgorski's tool, with input from Kaspar, Maedeh, Andrew, Manni, and others, focusing on receptive field mapping, noise correlations, and functional organization of dendritic inputs; the group discussed methodological details, limitations, and future directions.

Receptive Field Mapping Approach: Manav described the use of delta F/F traces from NWB files to map receptive fields, applying high spatial selectivity thresholds and filtering out unlocalized ROIs to ensure data quality; the analysis included pooling across mice and sessions to flatten out artifactual hotspots and focused on spatial selectivity in V1.

Cross-Validation and Artifact Control: Stefan and Kaspar raised concerns about the need for cross-validation to avoid artifacts in receptive field and noise correlation analyses; Manav explained the filtering and shuffling methods used, including session-based pooling and thresholding, but acknowledged the need for more rigorous cross-validation and controls, which were discussed as future improvements.

Retinotopic Gradient and Orientation Tuning: Manav reported that classical retinotopic gradients and strong relationships between receptive field and orientation tuning were not observed, with the analysis underpowered due to limited significant session planes; the group discussed the limitations of linear spatial fitting and the need for more sophisticated methods.

Noise Correlation Analysis: Manav explored noise correlations as a method to circumvent dendritic segmentation limitations, finding positive correlations across synapses and a strong relationship between physical proximity and noise correlation; further, receptive field overlap was shown to predict noise correlation, suggesting functional organization beyond topographic arrangement.

Controls and Methodological Limitations: Alexander and others questioned whether stimulus-driven responses confounded noise correlation findings, suggesting analysis of baseline correlations and spontaneous activity; Manav acknowledged these concerns and discussed shuffling and split-half controls, but agreed that further methodological refinement was needed, including analysis with static gratings and more robust controls.

Experience-Dependent Plasticity of Orientation Tuning: Ogulsuray presented their analysis on experience-dependent plasticity of orientation tuning in synaptic inputs to V1 neurons, focusing on the effects of oddball blocks and stimulus context, with input from Sarah, Jerome, Stefan, Alexander, and others, highlighting compartment-specific changes and methodological limitations.

Block-Based Comparison of Tuning: Ogulsuray compared orientation selectivity and tuning strength across different blocks (control and oddball) using polar plots and vector sum methods, analyzing both apical and proximal dendrites; the analysis showed variable changes in tuning strength and preferred orientation, with heterogeneity prompting population-level analysis.

Tuning Sharpness and Gaussian Fit: A constrained single Gaussian fit was applied to ROI tuning curves to measure tuning sharpness (half width at half height), revealing mostly unchanged and positively skewed distributions across blocks; Ogulsuray noted that tuning sharpness may be a fixed property and discussed possible modeling artifacts.

Population-Level Trends and Compartment Differences: Histograms of tuning strength and sharpness at the ROI population level within sessions showed small decreases in tuning strength and stable sharpness, with changes largely cancelling out at the mean; preferred orientation changes were more pronounced, especially in apical dendrites and superficial VIP somas, suggesting compartment-specific input sources.

Statistical Analysis and Limitations: Ogulsuray used Wilcoxon and Watson U-squared tests for statistical comparison, but small sample sizes limited significance; the study pooled imaging modalities and cell types due to subject numbers, and acknowledged that future analysis with more data could clarify trends and test directionality of changes.

Implications and Future Directions: The findings were consistent with previous studies showing dispersed afferent inputs with similar orientation preference, and Ogulsuray suggested that future work could explore whether within-block activity predicts changes in ROI responses; the group discussed the importance of these results and encouraged sharing of presentations for further input.

Discussion of Data Sharing and Continued Collaboration: Jerome, Kaspar, and others discussed sharing raw data and analysis repositories, encouraging Manav and Ogulsuray to post their presentations on the forum for ongoing feedback and collaborative refinement, with plans to continue methodological discussions and analysis improvements.

Data Access and Repository Sharing: Kaspar and Jerome noted that raw NWB files and analysis repositories are now available, enabling further analysis and methodological improvements; Kaspar offered to share papers and code to help Manav refine their approach.

Forum-Based Collaboration: Jerome encouraged both presenters to post their presentations on the forum, anticipating detailed feedback from the group and ongoing discussion of figures and statistical methods; the group agreed to continue conversations and anchor future work with shared raw data.

💬 Join the discussion for this page on GitHub