Limiting values (panel a) and decay times (panel b) of the washing function as a function of the initial probe intensity logIP(0): The dots are the probe-level data of all PM-probes of array A (see Eqs. (8) and (9)). The moving average was calculated over 1000 probe-level data to extract the mean effect of intensity on both parameters (thick curves). The moving average of the MM probes (probe-level data are omitted for clarity) is virtually indistinguishable from that of the PM probes. The PM-data are also split into probes which are hybridized predominantly non-specifically and specifically (thin lines, see text). The respective moving averages cover the low-intensity and high intensity ranges, respectively, with considerable overlap (see arrows in panel a). These results show that the washing parameters are mainly determined by the probe intensities and thus by the binding constant independently of the probe type (PM or MM) and of the hybridization mode (specific or non-specific). The mean trends of w∞ and τ are well described using Eqs. (18) and (17) given in the Methods-section (dashed lines). Accordingly, the stepwise change of the washing parameters is governed by their power law dependence on the binding constant. The fits use a critical exponent of γ = 1.6 and the critical intensities of logI(0)crit = 3.8 (for w∞) and 3.5 (for τ). The critical exponent and the critical intensity determine the sharpness of the sigmoidal change and the position of its inflection point, respectively (see also Figure 16 for illustration).