Extended Data Fig. 2: Resolving the “power gap” between single-target and ensemble holograms.

a, Illustration of the two compared stimulation conditions. An opsin-positive cell was patched and photocurrents were recorded in whole-cell configuration while the regular connectivity mapping protocol was used. Trials when single-target holograms were targeting the patched cell (left) were compared with trials when the patched cell was a part of a 10-target ensemble hologram (right). b, Photocurrents from an example experiment. Red stars indicate the photocurrent amplitude across increasing stimulation power for a single-target hologram applied to the patched cell, blue circles represent the response amplitude across power for different ensemble holograms containing the patched cell. c, Same data as in (b), but showing the mean ± s.e.m. for the different ensemble hologram sets (left), and normalized to the corresponding single-target photocurrents (right, magenta triangles). d, Mean photocurrents evoked by ensemble holograms, normalized to photocurrents evoked by single-target holograms at the corresponding laser powers. Error bars show one standard deviation. Data collected across a population of cells; n=6 experiments performed on 1 pyramidal cell (Emx-Cre), 1 PV cell (PV-Cre), 4 SST cells (SST-Cre). e-g, same as (b)-(d) but with laser power for ensemble holograms increased by 35%. For (g), data collected across n=2 experiments (1 pyramidal cell, 1 SST cell).