Fig. 2

Aging profiles characterized by stochastic variation. a Variance of the top 25,000 transcripts plotted using multi-dimensional scaling analysis; Y young, O old, n biological replicate. b Heatmap showing Euclidean distance matrix between young and old replicate samples calculated from the variance in count values where values closer to 1 are less variable. c, d The calculated mean squared coefficient of variation (CV²) for genes (b) and isoforms (d) plotted as a function of normalized count expression for young and old liver samples. e The percentage of normalized individual count data for each transcript type. Black bars represent the mean ± standard deviation