Fig. 5

Limit of Detection MA-plots. To assess the limit of detection for a given method (Life Technologies a-c; qpcR d-f), we plot the difference between the observed and expected average expression vs the expected expression for each feature in the two low input sample types (either the 0.1/0.1 or 0.01/0.01 dilution samples). The expected expression is calculated using either the pure sample expression (a, b, d, & e) or the 0.1/0.1 dilution expression (c & f). Each point represents an miRNA in the sample type of interest. Features for which all four data points are of good quality are shown in black; other features are shown in red with the proportion of poor quality values for each feature denoted by plotting symbol. One can visually assess the limit of detect by the expected expression value (x-axis) at which the difference between observed and expected expression begins to differ substantially from zero and the proportion of poor quality values increases