I've generated a new annotation file for you to use. It includes all TSSs, either annotated (named with the gene's main Flybase ID) or unannotated (named "RAMPAGEpeak_123456"). The corresponding expression table includes RPM expression values for all these peaks (same names). Concerning the expression file, there is some data processing involved. We do collapse PCR duplicates before computing expression. Then the normalization for depth for each time point is made to the total number of reads that fall into a peak (as opposed to all mappable reads, like in standard RNA-seq). Therefore, for each time point:
Peak_expression_RPM = Reads_in_peak / Sum_of_all_peaks * 1,000,000
The background correction for transcript abundance is only involved at the peak-calling stage, and thereby it affects the location and size of peaks, but it does not subsequently play a role in quantification.
I've included a separate file describing the 182 transposon-borne TSSs.
Best,
Philippe Batut
Gingeras lab
Watson School of Biological Sciences
Cold Spring Harbor Laboratory