Figure 1

Method implemented to correct the biases linked to transcript size. Size classes were built every 100 bp for transcripts < 5000 bp, as too few transcript numbers were observed with a size ≥ 5000 bp leading to scattered dots. The dotted line separates the red regression line corresponding to transcripts < 600 bp and the blue one to transcripts ≥ 600 bp. The vertical axis corresponds to log ₂ transformed values of read numbers. A) Cattle sample 1479. B) Drosophila sample SRR384925 (7132 genes with a single transcript). C) Human sample SRR1177729 (16,228 genes with a single transcript). D) Representation of sample 1479 after correction (green). E) RNA-Seq simulated data using rlsim, where Loess smoothing was applied to each series. The blue points correspond to a run where the sequenced fragments are in the range of 250–450 bp, the red one to the range of 450–650 bp, and the violet one to the range of 650–850 bp.