Fig. 3

Unselected libraries capture slightly less complex transcriptomes due to their reduced read depth. A Comparison of the number of unique genes identified (having 1 or more reads mapped) and their biotypes in each library. The right plot zooms into the “other” (orange) biotypes from the left plot. B Saturation analysis of the number of unique genes identified in each library. Each line indicates the mean of 100 repeated subsamples for each integer percentage of the overall number of reads in that library. Standard deviation of the mean at each subsample point was also plotted but is not visible at this scale. C Upset plot [8] showing the coordinance of genes identified in each library. The bars at the bottom left of the plot indicate the number of genes identified in each library (as shown in Subfigure A and B). The bars at the top of the plot indicate the number of unique genes identified by the subset of libraries indicated by the filled circle below. D Inverse CDF plots showing the fraction of unique genes having more than “X” read(s) in the X-axis library that were unidentified (having zero reads mapped) in the Y-axis library. The four comparisons shown each reached an unidentified fraction of zero at X = 16, 22, 42, and 17, respectively (indicating all of the genes in the X-axis library with more than “X” reads we identified in the Y-axis library)