Figure1: Inter-individual variation in whole blood gene expression signatures.

Whole blood was drawn from 75 healthy volunteers. Two volunteers donated samples on two occasions, resulting in a total of 77 samples. Genes with at least a 2.0-fold change in level of expression from the mean in at least 5 (of 77) samples are shown. The expression pattern of the corresponding ~370 genes is displayed in hierarchical cluster format where rows represent genes (unique cDNA elements), and columns represent experimental samples. Each expression measurement represents the ratio of fluorescence from the hybridized experimental sample to a reference sample. Missing or excluded data are represented by gray squares. Correlation coefficients were calculated between the gene expression of every gene and each parameter (neutrophil count, lymphocyte count, time of blood draw, age, and red cell distribution width) across the 77 samples. The correlation values are plotted as moving averages of 11 genes. Gene expression data were randomly permuted 10,000 times and the resulting correlation coefficients were compared with the actual data to derive a p-value. Dashed lines represent the coefficient above which all parameters were significantly associated with variation in expression of each gene (p<0.05). Supplemental data and enhanced versions of the figures, including searchable clusters and raw microarray data, can be found on this web site. (http://genome-www. stanford.edu/normalblood)