Fig. 7: Artificial theta oscillation in Tk-GAL4² ∩ Vglut neurons aversively biases an attractive visual bar.

a Time course of the experiment. Each fly underwent 10 trials with or without preceding air puffs in a randomized order. b Example time courses of a vertical bar’s angular position with or without prior air puffs. For each group, results of all trials of an example fly were pooled. Lines and shaded areas represent means and SEM, respectively. c Average bar fixation indices of “puff -“ trials and “puff +“ trials per fly. N = 50. In this and following, box plots are generated so that center line indicates median, box limits indicate upper and lower quartiles, and whiskers indicate 1.5x interquartile range. **p < 0.01, *p < 0.05, n.s.: p > 0.05, two-tailed t-test followed by Bonferroni correction for each puff condition, and two-tailed paired t-test for between-conditions comparison. d Average wavelet cross-spectrum of the neural recordings. Red rectangle indicates the area of the θ frequency band which was observed when a small object was presented (Fig. 6b). N = 98 cells from 7 flies. e Power of θ oscillatory activity subtracted by power of the corresponding band during the 0.5 s time bin immediately prior to each time window, for each cell. N = 98 cells from 7 flies. ***p < 0.001, *p < 0.05, n.s.: p > 0.05. Two-tailed t-test followed by Benjamini–Hochberg correction for each condition, two-tailed t-test followed by Bonferroni correction for between-conditions comparisons. Dots and error bars represent means and SEM. f A schematic of hypothesis. g Example time courses of the visual object’s angular position. For each photoactivation frequency, results of all trials of an example fly were pooled. Lines and shaded areas represent means and SEM, respectively. h Light-induced changes in the bar fixation indices. N = 21, 22, for retinal-unfed and retinal-fed groups, respectively. **p < 0.01, n.s.: p > 0.05, two-tailed t-test or Wilcoxon signed rank test followed by Benjamini–Hochberg correction.