Figure 2From: Probe-level linear model fitting and mixture modeling results in high accuracy detection of differential gene expressionMixture modeling of the PL-LM method on the Choe et al. dataset. The three mixture components are represented as ellipse identifying their center and variances. Data points with a conditional probability of belonging to the component modeling DEGs above 0.5 are shown as larger gray dots. The mixing proportions of the three components are 17%, 74% and 9%, respectively for the non-spiked probe-sets, not amplified cRNAs and DEGs.Back to article page