A scaling normalization method for differential expression analysis of RNA-seq data

Mark D. Robinson [1,2] and Alicia Oshlack [1]

[1] Bioinformatics Division, WEHI
[2] Epigenetics Laboratory, Garvan
Corresponding author: mrobinson@WEHI.EDU.AU and oshlack@WEHI.EDU.AU

Summary

The fine detail provided by sequencing-based transcriptome surveys suggests that RNA-seq is likely to become the platform of choice for interrogating steady state RNA. In order to discover biologically important changes in expression, we show that normalization continues to be an essential step in the analysis. We outline a simple and effective method for performing normalization and show dramatically improved results for inferring differential expression (DE) in simulated and publicly available data sets.

Supplementary Data:

• Data used
  • (Figure 1) data from Marioni et al. [LK_data.RData]
  • (Supplementary Figure 3) data from Cloonan et al. [Grimmond_lengths.txt]
  • (Supplementary Figure 4) data from Li et al. remapped [li_mapped.tar.gz]
  • Human housekeeping genes [human_housekeeping.txt] - from Eisenberg and Levanon, 2003
  • Mouse housekeeping genes [mouse_housekeeping.txt] - from de Jonge et al., 2007
• R scripts
  • accessory R functions for simulation, analysis, plotting [functions.R]
  • analysis of Marioni and simulated data for Figures 1+2 [figures_1,2.R]
  • simulation comparing methods [figure_3.R]
  • code to generate supplementary figures [figures_supp.R]