The Immune Variation (ImmVar) project assayed gene expression in CD14 +CD16 − monocytes and CD4 + T-cells on the Affymetrix Human Gene 1.0 ST Array platform in order to characterize the role of cell type in genetic regulation of gene expression [1]. Analysis of 398 individuals with data from both cell types reveals that multiple variables contribute to expression variation in this dataset (Fig. 4 a). Since variancePartition reports the contribution of each variable while simultaneously correcting for all other values, it is apparent that the variation across cell types is the strongest biological driver of variation (16.4%) followed by variation across individuals (5.6%). Although cell type has a smaller median effect than batch, it is notable that cell type explains >50% of the variation for 4,591 genes. The observation that batch and cell type are the strongest drivers of variation is largely consistent with results from principal components analysis (PCA) (Fig. 4 b). We note that the relationship between variancePartition and PCA depends on both the fraction of expression variation explained by a particular variable across all genes as
