Finally, to compare information results between animal populations (P vs Wistar), we used a bootstrap approach to compare the weighted mean MI between P and Wistar rats at each time bin. We compared the difference between the weighted mean MI values in the real data to the difference weighted mean MI results from 10,000 randomized trials (identity of P and Wistar neurons randomized preserving number of neurons in each group). The p value was then calculated as the proportion of randomized trials with differences greater than or equal to the difference in the real data, accounting for the sign of the difference. In the case where all randomized trial difference values were less than the result from the real data, the p value was set to 0.00005 = 0.5 × (1/10,000) due to the resolution associated with using 10,000 randomized trials. These p values were then corrected for multiple comparisons across time bins within a given figure using FDR control (Benjamini and Hochberg, 1995).
