Fig. 3

Technical replicate. Twelve replicate transcript counting libraries were prepared from 1 μg of a pool of wild-type zebrafish embryo total RNA sample. Libraries were sequenced by Illumina MiSeq and analysed using the DeTCT pipeline. a The normalised counts for each region were extracted (73,938 regions). After filtering the data for only the regions which we would use to call differential transcript abundance the counts from all 12 libraries were compared using a Pearson correlation (see Additional file 5). Cells coloured yellow in the Pearson correlation are the most highly correlated while those in blue are the least correlated with a colour gradient inbetween. b In addition four concentrations of ERCC spike mix 1 were added in triplicate to the same 12 libraries prior to library construction. We added the quantity suggested by the manufacturer (x1), five times the quantity (x5), one fifth of the quantity (x0.2) and one tenth of the quantity (x0.1). The reads were mapped to the zebrafish reference sequence and ERCC spike reference sequence. The diagram shows the 92 ERCC spikes represented by a circle in descending order of spike copy number in the mix on the x-axis and spike abundance on the y-axis. The blue circles show spikes identified in the DeTCT pipeline while those in red were not found. c The DeTCT analysis pipeline was run using six libraries at a time with three replicates as one pair of conditions and in all six possible condition combinations. The mean log₂ fold change was calculated for all the spikes detected by the DeTCT analysis and plotted against the expected log₂ fold change as circles. Each circle is labelled with the conditions being compared and observed log₂ fold change over the expected log₂ fold change. The numbers in brackets indicate how many spikes show differential transcript abundance