Fig. 4: Sources of variation from the experimental processes.
From: A real-world multi-center RNA-seq benchmarking study using the Quartet and MAQC reference materials
Scatterplots of PCA on RNA-seq data from 26 laboratories for the (a) Quartet samples, and (b) MAQC samples after applying the fixed analysis pipeline. The circles of the same color represent all replicates across all laboratories for each sample. Principal variance component analysis quantifies the proportion of variance explained by each experimental factor for the (c) Quartet samples, and (d) MAQC samples. Heatmap and hierarchical clustering of laboratories based on the root mean square error (RMSE) at (e) absolute expression levels, and (f) relative expression levels for the Quartet samples. The average gene expression values across three replicates of the four Quartet samples for the intersecting genes between each laboratory’s data and the Quartet reference datasets (n = 30,956) were used to calculate the RMSE between any two laboratories or between a laboratory and the Quartet reference datasets. The box plots present the RMSE values for 26 laboratories, and data are presented as median values (center lines) and the upper and lower quartiles (box limits).
