The Sequencing Quality Control (SEQC) project [9] evaluated the technical reproducibility of RNA-seq data by sequencing the same 4 RNA samples at 6 laboratories, using 108 total library constructions and up to 8 lanes on each of 11 Illumina HiSeq 2000 flowcells for a total of 1580 RNA-seq experiments. The goal of the study was to determine the degree to which these technical factors explain variation in gene expression measurements. This complex dataset has multiple levels of variation and variancePartition provides a rigorous statistical framework to quantify and interpret these sources of variation in a single analysis.
