Skip to main content
Fig. 3 | BMC Systems Biology

Fig. 3

From: DGCA: A comprehensive R package for Differential Gene Correlation Analysis

Fig. 3

Definition of differential correlation classes. This diagram demonstrates the definition of differential correlation classes used throughout DGCA in the case of two conditions. The default p-value significance level threshold (α) parameter setting for the hypothesis test of non-zero correlation in one individual condition in DGCA and throughout this manuscript is 0.05, although users of DGCA can adjust this parameter if they choose to. In each condition, gene pairs are defined as having a non-significant correlation (p-value < α), a significant and positive correlation (p-value < α, correlation (ρ) > 0), or a significant and negative correlation (p-value < α, correlation (ρ) < 0). The Cartesian product of the 3 possible correlation classes in one condition with the 3 in the other condition yields a total of 9 possible differential correlation classes. Note that, theoretically, a gene pair with a significant positive correlation in one condition and a non-significant correlation in another condition may not be significantly differentially correlated between these conditions since the correlation class identification is independent of the differential correlation hypothesis test

Back to article page