Fig. 3

\(3^\prime\) Normalization estimates depend on assumed polymerase elongation behavior and sequencing depth A A cartoon showing the characteristic shape of nascent RNA sequencing samples after a perturbation. RNA polymerase II loads at the 5\(^\prime\) end of genes, thus after a perturbation alterations in transcription levels can only reach a distance that depends on the processivity of RNA polymerase II. In this work we assume 3kb/min and hence for a 60 min experiment the perturbation influences the first 180kb (\(60\text {min}\cdot 3\frac{{\text {kb}}}{{\text {min}}}\)). B We compared external spike-ins (y-axis) to 3\(^\prime\) internal normalization across a large collection of previously published data. Samples are colored by whether they C meet both time point and depth assumptions (green), D have low sequencing depth (\(<10\)X spike-in transcriptome) (orange), E have time points beyond the \(3^\prime\) assumed 60 min (blue), or F meet neither assumption, being of both low spike-in depth and long time point (red). Notably, two samples in (circled in C) meet the coverage and time constraints of the \(3^\prime\) normalization approach but involve depletion of NELF under heat shock conditions, which likely alters RNA polymerase elongation