Synthetic spike-in standards for RNA-seq experiments

• DOI: 10.1101/gr.121095.111
• Publication Date: 2011-08-05T02:15:35Z
• Authors: Lichun Jiang; Felix Schlesinger; Carrie A. Davis; Yu Zhang; Renhua Li; Marc Salit; Thomas R. Gingeras; Brian Oliver
• Keywords: RNA-Seq

High-throughput sequencing of cDNA (RNA-seq) is a widely deployed transcriptome profiling and annotation technique, but questions about the performance different protocols platforms remain. We used newly developed pool 96 synthetic RNAs with various lengths, GC content covering 2 20 concentration range as spike-in controls to measure sensitivity, accuracy, biases in RNA-seq experiments well derive standard curves for quantifying abundance transcripts. observed linearity between read density RNA input over entire detection excellent agreement replicates, we significantly larger imprecision than expected under pure Poisson sampling errors. use control directly reproducible protocol-dependent due transcript length stereotypic heterogeneity coverage across transcripts correlated position relative termini priming sequence bias. These effects lead biased quantification short individual exons, which serious problem measurements isoform abundances, that can partially be corrected using appropriate models By RNAs, limits discovery rare experiments. data collected part model organism human Encyclopedia DNA Elements projects (ENCODE modENCODE), demonstrate external are useful resource evaluating sensitivity accuracy quantification. quality metrics facilitate comparable analysis samples, protocols, platforms.