Figure 2
From: Reducing bias in RNA sequencing data: a novel approach to compute counts
Exon length bias and GC-content effect. Smoothed scatter-plots showing the relationship between log-counts/RPKMs and exon length (log scale, A) or GC-content (B), in Jiang's data ("nucleus" libraries). The orange curve represents a cubic-spline fit computed on log-counts, averaged in bins of 5000 exons each (black crosses between vertical lines, indicating bin bounds). Counts or RPKMs are computed using totcounts, maxcounts, RPKM-corrected totcounts (RPKM) and totcounts corrected with within-lane full-quantile normalization over exon length (FullQ), and averaged across libraries.