Orientation selectivity of late V1 response.

<p>(A) Left, raw traces in 10 trials of cell-attached recordings, raster plots of spike responses in 80 trials, and periflash time histograms of the firing rates for four orientations of the grating flash stimulation in a representative neuron (<a href="http://www.plosbiology.org/article/info:doi/10.1371/journal.pbio.1002231#pbio.1002231.s003" target="_blank">S3 Data</a>). The orientations are shown in different colors. Right, the orientation tuning curve of the same neuron. The evoked spike counts were normalized to the maximum. (B) The cumulative probability distribution of the OSIs of the 36 late-spiking cells (Real) was compared with its chance distribution (Surrogate) that was obtained by 1,000 random shufflings of the stimulus trials. The real OSIs were biased rightward compared with the surrogate OSIs (<i>p</i> = 3.3 × 10⁻³, <i>D</i> = 0.29, Kolmogorov-Smirnov test). (C) Left traces represent the mean ± SD of subthreshold <i>V</i>_<i>m</i> responses of an example cell (<a href="http://www.plosbiology.org/article/info:doi/10.1371/journal.pbio.1002231#pbio.1002231.s003" target="_blank">S3 Data</a>) to grating flashes with four orientations. The right plot indicates the orientation tuning curves of the mean amplitude of the early and late <i>V</i>_<i>m</i> depolarizations of the same neuron. (D) The cumulative fraction of the OSIs in the late <i>V</i>_<i>m</i> responses were biased rightward compared with their surrogate OSIs (<i>n</i> = 34 cells, <i>p</i> = 2.7 × 10⁻⁹, <i>D</i> = 0.66, Kolmogorov-Smirnov test). (E) The correlation coefficients between the early and late tuning curves for individual cells were higher compared with their chance values calculated by random trial-shuffling of the early responses (<i>n</i> = 34 cells, <i>p</i> = 0.014, <i>D</i> = 0.27, Kolmogorov-Smirnov test). (F) Scatter plots of the OSIs in early and late responses for individual cells. Each dot indicates a single cell. The gray line is the diagonal, and the dash line is the best linear fit.