Figure 4

Disease genes show distinct topological features even when considering function. Q-Q plots (left-most columns) and difference plots (right-most columns) for degree (a), clustering coefficient (b) and betweenness centrality (c). The Q-Q plots are obtained by plotting the quantile of the samples against the quantile of the disease sets, in the 25th - 75th percentile range. Function-constrained samples are shown in red, unconstrained samples in blue. Note that the samples are compared to the disease sets to make the Q-Q plots. The x-coordinate of a point in the Q-Q plot is the value of the given topological property in the disease set at a given percentile, whereas its y-coordinate is the average value of the topological property in the samples at the same percentile. Difference plots show the difference between the topological property of a disease set at a given quantile and the topological property of the samples at the same quantile. 1000 function-constrained and 1000 unconstrained samples have been generated for each disease set. The shaded areas encompass the interval between the bottom 5% and the top 95% of the values at a given quantile. See Additional file 1 Table S3 for empirical p-values assessing differences between topological features in disease and sample sets at the 25th, 50th and 75th percentiles.