Figure 8

Geometrical and topological network similarity. Two Keiko maps are compared, one constructed on MG-U74v2-NR1 and another on RG-U34-NR1. Only unique and multiple probe set pairs are taken into account in the display and statistics. A – Six clusters found by Gene DIVER in the RG-U34-NR1 map are displayed in the upper right corner. The corresponding points on the MG-U74v2-NR1 map are displayed in the lower left corner. All the points belonging to a given cluster in the RG-U34-NR1 map are linked by a coloured line, and the same is true for the corresponding points in the MG-U74v2-NR1 map. B – Two points that correspond to a pair of probe sets are linked by a line coloured according to the cluster they belong to. C – In blue is plotted the probability density of the distance in the RG-U34-NR1 map between one point and the closest point in the MG-U74v2-NR1 map. In red is plotted the probability density of the mean distance in a series of random permutations of RG-U34-NR1. The black vertical line indicates the median value of the first distribution. The magenta vertical line marks the mean of the 100 median values calculated on permuted maps. The ratio between these two values is used to score the "distance similarity" between the two maps. D – In blue is plotted the probability density of the ranks in the RG-U34-NR1 map between one point and the closest point in the MG-U74v2-NR1 map. In red is plotted the probability density of the mean rank in a series of random permutations of RG-U34-NR1. The ratio between the two values marked by the two vertical lines is used to calculate a ranking score that scores the "rank similarity" between the two maps.