Figure 19
Scatter plots of residuals for mac1 versus log2(Cy3) as independent variables together with various STIs. (a) Residuals for artificial data drawn from a bivariate normal distribution with the same parameters as the real mac1 data shown below. This plot is based on the same simulated dataset as in Figure 17b except that the linear trend has been subtracted resulting in slope = 0. A robust scale estimator (a Huber τ-estimate of scale) for residuals (green, outer), supsmu (purple, middle) and lowess-based (orange, inner) scale estimators are shown. Scale factors calculated to adjust supsmu- and lowess-based scale estimates were 1.25 and 1.35, respectively. (b) Adjusted supsmu-based STIs for the real data at the 95% (innermost), 99%, 99.8%, 99.98% and 99.998% (outermost) levels. These adjusted STIs take into account differences between the ordinary STIs and supsmu-based STIs for an artificial dataset having the same parameters (location, scale and coefficient of correlation) as its cognate real data. The vertical dotted line marks the location of the minima of the empirical hyperbolas. Red or pink and blue or cyan dots correspond to up-regulated and down-regulated genes, respectively, with p-value ≤ 0.01 (≤ 0.05). The most prominent down-regulated gene with case index number 4,301 is Mac1 (see also Table 3), which is not expressed in the mac1 strain.