What is the significance of the r-value, p-value and E-values?
We have tried to give an indication of the biological significance of any r-value by calculating a p-value and an E-value for each result. For an observed correlation ro, the p value was defined as p = P(r >= ro) if ro >= 0, or p= P(r <= ro) if ro <= 0, using bivariate normal distribution, representing the probability of observing an equal or larger positive or negative correlation by chance. The E value was defined as the expected frequency of such chance observations on an array of M genes, E=Mp, and is directly analogous to the E values reported by the popular sequence search program FASTA (Pearson and Lipman, 1988). The E value is a convenient way of accounting for multiple testing effects in whole genome studies. This procedure is related to the standard Bonferroni correction for multiple sampling. There is certainly no single r-value that can be used as a cut-off either in one correlation list or generally.
