Fig. 2

Impact of probe intensity on β value reliability in the TACT study. a Association of probe variability within participants at visit 1 (SDβ) and over time (∆β) with mean intensity (MI). Probes with low MI have high variability over time. Examples of a probe with low MI (b) or high MI (c). Lines are connected by points corresponding to the individual in two different visits. β values from the same individual are closer for a probe with a high MI than on a probe with a low MI. d Reliability map of β values. e Smoothed curve shows the dependence of type II probe unreliability on mean intensity (MI). The point highlighted on the graph is the “critical point” marking a sharp change in dependence decline, which we recommend as a dynamic threshold for determining which probes are deemed unreliable (for type I probes and other datasets see Supplementary Figs. S4–S6). f Dependence of the type II probe unreliability on MI, highlighting probes which are detected using the p-value method at different threshold stringency (for type I probes and other datasets see Supplementary Fig. S12)