Extended Data Fig. 3: Changes in baseline decoding accuracy following preconditioning and aversive learning. | Nature

Extended Data Fig. 3: Changes in baseline decoding accuracy following preconditioning and aversive learning.

From: Prefrontal encoding of an internal model for emotional inference

Extended Data Fig. 3

(a) Peri-event time histograms showing population averaged auditory evoked calcium responses (lines) and SEM (shaded) of activated (solid) and suppressed (dashed) neurons without baseline subtraction during preconditioning (left) and recall (right) for paired group. (b, c) Top: Peri-event time histograms showing decoding accuracy for distinguishing auditory stimuli from baseline activity based on dmPFC population activity (including all neurons) during preconditioning (PC, b) and recall (c) in the paired (‘P’, red, solid) and unpaired (‘unP’, grey, dashed) groups as well as corresponding shuffled paired (‘PS’) and unpaired (‘unPS’) control datasets. Bottom: Box plots showing median (center), 25%/75% interquartile range (bounds of box) and extreme data points (whiskers) of decoding accuracy for auditory stimuli from the peri-event time histograms in the top panels. Kruskal-Wallis test (two-sided), *: X2 = 11, p = 0.012; ***: X2 = 25.72, p = 1.1 × 10−5; Post-hoc test, p(a,unP-PS) = 0.012; p(b,P-unP) = 0.0013; p(b,P-PS) = 0.0013; p(b,P-unPS) = 1.5 × 10−5; (d, e) Similar to b, c, but for visual CSs during preconditioning (d) and recall (e). Kruskal-Wallis test (two-sided), n.s.: X2 = 4.76, p = 0.19; ***: X2 = 26.85, p = 6.3 × 10−6; Post-hoc test, p(d,P-PS) = 1.6 × 10−4; p(d,unP-PS) = 0.0028; p(d,P-unPS)=0.001; p(d,unP-unPS)=0.013; (f) Box plots showing median (center), 25%/75% interquartile range (bounds of box) and extreme data points (whiskers) for decoding accuracy (auditory response vs baseline) during preconditioning and recall for the paired group. Mann-Whitney U test; *: z = 2.30, r = 1285, p = 0.022. (g) Similar to f, but for decoding accuracy in the unpaired group. Mann-Whitney U test; *: z = 2.14, r = 1273, p = 0.032. (h) Decoding accuracy (visual response vs baseline) during preconditioning and recall in the paired group. Mann-Whitney U test (two-sided); ***: z = 4.11, r = 1426.5, p = 3.9 × 10−5. (i) Similar to h, but for decoding accuracy in the unpaired group. Mann-Whitney U test (two-sided); **: z = 2.64, r = 1312, p = 0.0082. (j) Auditory (red) and visual (blue) stimulus evoked calcium responses of three example neurons during memory recall. Each trace shows responses on a single trial. (k, l) Spearman correlations between averaged calcium responses in single neurons in response to auditory and visual stimuli for the paired (left) and unpaired (right) groups during preconditioning (k) and recall (l). Auditory-only cells (red), visual-only cells (blue) and co-responsive cells (purple) are denoted by different colors. Each dot represents one neuron. (m) The distribution of correlation coefficients in the unpaired group between auditory and visual CS-evoked calcium responses during preconditioning by bootstrap resampling. The correlation coefficient in the paired group (red line) is within the 95% confidence interval (dashed lines). (n) Similar to m, but for the correlation coefficients of neural response during recall which falls outside of the 95% confidence interval.

Source Data

Back to article page