E-MTAB-5446 - RNA-seq of wild type Arabidopsis seedlings to compare cross-replicate variability and false discovery rate in differential gene expression tools
Submitted on 26 May 2014, last updated on 8 February 2019, released on 8 February 2019
RNA-Seq data from 17 wild-type biological replicates of Arabidopsis thaliana used to explore read count measurements across replicates along with the False Discovery Rate of Differential Gene Expression tools. Although A. thaliana has a relatively small genome, its transcriptome is similar in scale and complexity to that of model mammal species and its genome is extensively annotated and the conclusions presented here provide useful guidance for work in other complex eukaryotes. The findings show that the negative binomial and log-normal distributions are both good choices as models for the cross-replicate variability of RNA-seq read counts. 6 of 9 DGE tools controlled their identification of false positives well even with only 3 replicates. Our results reinforce the conclusions reached by Schurch et. al. (2015 RNA) in yeast.
RNA-seq of coding RNA, RNA-seq of non coding RNA, genotype design, replicate design, self vs self design, software variation design