By applying a conservative threshold of <0.4, we identified voxels that either are strongly orientation selective or have no selectivity (responded to all orientations).

Using 1 − circular variance to examine preferred orientation on a voxel-by-voxel basis reveals clustering of voxels according to motion preference as derived from the calculated complex angle (Figure 3B). Polar plots from individual voxels illustrate such highly orientation-selective selleck chemicals responses (Figure 3C). Examining the cumulative distribution of complex angles across all larvae imaged reveals two clear populations (Figure 3D) plus a baseline component that reflects the voxels responding to all orientations with noise randomly and evenly distributing the calculated complex angles. Iteratively fitting two summed von-Mises distributions (constrained with bimodal distributions separated JQ1 purchase by 180° and equal concentration) plus a baseline component to the histogram data derived distinct population peaks centered at 105°/285° and 172°/352° (Figure 3D).

These correspond to motion of vertically oriented bars moving along the horizontal axis (horizontally tuned) and horizontally oriented bars moving along the vertical axis (vertically tuned), respectively (Figure 3G). The largest fraction of orientation-selective voxels is tuned to vertical motion. Within all individual larvae examined, the relative proportions of voxels selective for vertical and horizontal motion generally reflect those in cumulative population data. From the distributions identified in Figure 3D, we generated parametric maps in which voxels are color coded according to orientation preference and superimposed on the fluorescence however image of the tectal neuropil (Figures 3E and 3F). These maps, which allow examination of functional architecture in individual larvae, reveal that in all subjects, orientation-selective inputs are broadly distributed across SFGS

and that voxels tend to cluster according to orientation preference. What is evident from the two examples of separate larvae is that within the orientation-selective domain, the organization of the two subtypes can be variable across subjects. The same orientation-selective inputs, with similar tectal distributions, were identified using the OSI metric (Figure S3). This figure also shows examples of single orientation-selective RGCs expressing SyGCaMP3 that are selective for either horizontal or vertical motion. The functional parametric maps of individual larvae shown in Figures 2 and 3 suggest regional differences in the distribution of direction- and orientation-selective inputs to the zebrafish tectum. To examine in more detail the spatial organization of direction- and orientation-selective responses in SFGS, we spatially coregistered data from all larvae to create single composite maps for each parameter (see Supplemental Experimental Procedures).