An example of a co-occurrence matrix. (a) A binding profile for two proteins, A and B, are displayed schematically in red and blue respectively. The sequence coordinate is along the horizontal axis and each rectangle represents a location. (b) A schematic of the co-occurrence matrix for the joint profile of (a). Locations are numbered sequentially by start coordinate. Rows and columns are associated with factors A and B respectively. The metric is used is whether locations overlap. Overlapped pairs have non-zero c
and are indicated by diagonal hatching of the element. For example, we see from the matrix that locations 3 and 4 overlap and when they are bound by factors A and B, c34 is non-zero and therefore the statistic is increased by its value. (c) The binding profiles of A and B shown in (a) have been indicated on the co-occurrence matrix with faint red and blue backgrounds respectively. The elements that will be summed in evaluating the statistic are outlined in bold. These represent every pairwise combination of the two profiles.