Supplementary Figure 3: ΔAzimuth is larger than ΔElevation for both LM boutons and L2/3 somata. | Nature Neuroscience

Supplementary Figure 3: ΔAzimuth is larger than ΔElevation for both LM boutons and L2/3 somata.

From: The functional organization of cortical feedback inputs to primary visual cortex

Supplementary Figure 3

a, ΔRF of V1 L2/3 neurons measured with a square stimulus grid (azimuth in [−30°, 30°], elevation in [−30°, 30°]). b, Distribution of relative retinotopic position (left, ΔAzimuth; right, ΔElevation) of L2/3 V1 neurons. c, Cumulative distribution of ΔAzimuth (blue) and ΔElevation (red) distances. **, P = 1.6x10-22, two-sample Kolmogorov-Smirnov test, two-sided, n = 1979 L2/3 neurons. d-f, same as in a, b, and c but for LM boutons. **, P = 1.3x10-16, two-sample Kolmogorov-Smirnov test, two-sided. n = 3423 LM boutons. g, Angular histogram of deviation angle θ of LM boutons in d. Boutons less than 10° away from the origin are discarded for angular counts. Inner circle, expected fraction of a uniform distribution (0.125). Boutons are enriched in the horizontal bins even when RFs are measured using a square stimulus grid. h, Same as g but only for gratings responsive, non-selective boutons. Bottom, normalized population tuning curve. The group is slightly more tuned for horizontally-moving vertical stimuli. i, Same as h, after removing the most tuned boutons until tuning for horizontal and vertically-moving stimuli was equal (bottom). j, Cumulative distribution of ΔAzimuth (blue) and ΔElevation (red) distances for gratings responsive, non-selective boutons after removing the most tuned boutons for horizontally-moving vertical stimuli. *, P = 0.048, two sample Kolmogorov-Smirnov test, two-sided, n = 146 boutons.

Back to article page