Fig. 3

(A) Comparison of the chromatic computations occurring between AF10 and tectum. (B) Anatomical layout of AF10 (top) and ipsilateral tectum (bottom). Scale bar, 100 μm. (C) Histogram quantifying the number of ROIs assigned to the top 75th percentile of each stimulus, no stimulus (NR, non-responsive), or multiple stimuli for both populations. p > 0.05 for all bins tested separately, Wilcoxon signed rank. (D) Overall decorrelation of the neural activity evoked by each stimulus for both datasets. (E) Cluster average decorrelation for all pairs of stimuli for AF10 (below diagonal) and tectum (above diagonal). The deltas were evaluated pairwise; p < 0.05, Wilcoxon rank sum for all squares. (F) CCA-aligned and averaged PC trajectories over the course of the stimulation period for AF10 and tectum. Line color indicates stimulus, with dot size increasing over time. (G) Average pairwise Euclidean distances between the centroids of the point clouds corresponding to each stimulus response in PC space for AF10 and tectum after alignment. ∗p < 0.05, Wilcoxon rank sum performed pairwise. (H) Example SVM-based classification performance over time for AF10 and tectum compared to shuffle. Stimulus is on top. (I) Success ratio for classification of each stimulus in AF10 compared to tectum. ∗p < 0.05, Wilcoxon rank sum. (J) Classification performance for different ROI numbers, with an average of 10 repetitions for both populations and shuffles. * p < 0.05 Wilcoxon rank sum performed pairwise. (A–E, H, and I) n = 6 animals and 1,776 ROIs for tectum; n = 6 animals and 2,670 ROIs for AF10. (F and G) n = 3 animals and 1,604 ROIs for tectum; n = 5 animals and 1,313 ROIs for AF10.