Figure 6

Data set comparison. Eleven microarray data sets (A-K) exhibiting varying degrees of conformation to power law and log normal distributions. On the left for each data set is a log mean intensity vs. log rank plot of the entire data set. Each array was sorted independently by intensity, and mean intensities for each rank over all arrays are plotted. A linear regression line is shown in red. Data sets with a linear distribution adhere well to a power law distribution. On the right for each data set is the distribution (ln(i) – μ) / σ of the mean intensities used in the left hand plots, where i is the mean measured intensity for each rank and μ and σ are the mean and variance of i respectively. The standard normal curve N(0,1) is shown in red for comparison. Data sets that display a standard normal distribution adhere well to a log normal distribution.