PARE experimental procedure and peak finding analysis pipeline.

A) Diagram of the PARE procedure. B) Schematic of PyDegradome analysis approach, which uses read counts in a control sample to generate a table of thresholds to compare test sample counts to. The table lists thresholds for a particular user-defined confidence level and for a range of ratios between control and test samples. The applicable ratio for each position is computed by multiplying a user-defined multiplicative factor by the ratio of total read counts for the exon in test vs. control samples, thus accounting for variation in RNA levels and total mapped reads. Read counts in test sample at each position must exceed the threshold to be identified as part of a peak. C) Example of plot of read counts (5’ end only) for 200 nt surrounding a SOX cut site identified by PyDegradome within the 3’–most exon of the LIMD1 NM_014240 transcript in the four samples. Note that y-axis has a logarithmic scale. This example shows the expected distribution for a cut site followed by exonucleolytic degradation due to residual Xrn1 activity or to the action of another nuclease, with the highest read count at cut site and decaying counts at following positions. Positions referred to as “peak” and “cut site” in the text are marked on the graph. D) Flow chart of steps used to defined SOX cut sites used for further analyses.