Journal: bioRxiv

Article Title: Top-down feedback enables flexible coding strategies in olfactory cortex

doi: 10.1101/2021.08.06.455459

Figure Lengend Snippet: (A). Firings of piriform cells (PCs) in response to a representative odor with centrifugal feedback OFF. PCs fire with a transient burst of spikes that are sharply truncated by local FBIs and followed by persistent suppression. Population firing rates: (mean ± SD, n = 10 trials). (B). Similar to (A) but for centrifugal feedback ON. Persistent dynamics of PCs arise due to centrifugal feedback. FBI cells are only sparsely activated. (C). PC population firing rate in response to all model odors (mean ± SD, n = 300 odors). The traces of feedback OFF and ON are normalized to have the same amplitude of peak. (D). Schematic illustration to quantify the dynamics of the PC population firing rate in response to a single odor ( n = 300 odors). Peak: the first peak in the trial-averaged population firing rate ( n = 10 trials). Slope: the slope of a linear function (oblique dashed line) fitted to the mean firing rate between the Peak and the first time when it drops below baseline (firing rate preceding the activation of the earliest glomerulus). Delay: the latency between the Peak and the activation time of earliest glomerulus defined by the odor (vertical dashed line with a triangle on top). (E). Comparison of PC dynamics between centrifugal feedback OFF vs. ON. (E1). Peak firing rate. Error bar: ±SD (*** p < 0.001 Wilcoxon signed rank test). Connecting lines for 30 example odors are shown. (E2). Similar to E1 but for Slope. (E3). Histogram of the Delay across 300 odors. Top: M/T cells; bottom: PCs. Centrifugal feedback reduces the responses latency of PCs without affecting that of M/T cells.

Article Snippet: Statistical tests for significance were performed with a two-sided Wilcoxon rank sum test (ranksum function in MATLAB) when samples were independent ( and ) and with a two-sided Wilcoxon signed rank test (signrank function in MATLAB) for paired samples ( and ).

Techniques: Activation Assay

Journal: bioRxiv

Article Title: Top-down feedback enables flexible coding strategies in olfactory cortex

doi: 10.1101/2021.08.06.455459

Figure Lengend Snippet: (A). Piriform cell (PC) responses to Odor-1 and Odor-2 with centrifugal feedback OFF. Top and middle: raster plot of piriform cell responses to Odor-1 and Odor-2 in an example trial. Bottom: population firing rates of PCs responding to two odors (mean ± SD, n = 10 trials). The firing rate separations between Odor-1 and Odor-2 are nonsignificant (ns: p > 0.05 Wilcoxon rank-sum test). (B). Similar to (A) but for centrifugal feedback ON (mean ±SD, n = 10 trials). The firings of PCs are persistent throughout successive glomerular activation of both odors. The firing rate between Odor-1 and Odor-2 are significant different during 70 – 90 ms and 120 – 140 ms (** p < 0.01 Wilcoxon rank-sum test). (C). Low-dimensional projections of ensemble trajectories of PCs onto the first three principal components when centrifugal feedback is OFF. Each trace corresponds to a single-trial PC response to one of the odors (color coded). Trajectories for different odors are tangled and non-separable. (D). Similar to (C) but for centrifugal feedback ON. Centrifugal feedback pushes apart the low-dimensional trajectories evoked by different odors and makes them more separable.

Article Snippet: Statistical tests for significance were performed with a two-sided Wilcoxon rank sum test (ranksum function in MATLAB) when samples were independent ( and ) and with a two-sided Wilcoxon signed rank test (signrank function in MATLAB) for paired samples ( and ).

Techniques: Activation Assay