Skip to main content
Figure 3 | BMC Systems Biology

Figure 3

From: Disentangling function from topology to infer the network properties of disease genes

Figure 3

Function-constrained samples closely match the functional composition and number of terms per protein of disease sets. (a) Boxplots show the Euclidean norm between the distribution of informative term functional annotations in the disease set versus those found in the samples; this corresponds to the square root of the functional distance defined in the Methods. Function-constrained samples are shown in the left columns and the unconstrained samples are shown in the right columns. The functional distribution of the function-constrained samples are close to the target distribution in all samples. (b) Distribution of the average number of terms per protein in the function-constrained samples (left columns) and unconstrained samples (right columns, blue). The average number of terms in each disease set is shown as a horizontal green line. Function-constrained samples have similar average number of terms per protein as the disease sets, whereas the unconstrained samples differ substantially.

Back to article page