Meta-Analysis of Differentiating Mouse Embryonic Stem Cell Gene Expression Kinetics Reveals Early Change of a Small Gene Set
Figure 2
Transcriptome Plots of Estimated Expression Changes, Based on Fitting Models to Each Dataset
All plots have density shading to demonstrate the number of points (genes) in different regions. Lines illustrate examples of some of the requirements that make up the definition of the ESC signature change and observed gene expression patterns that fulfilled all requirements are marked as ♦. Experiment-specific implementations of requirements are explained below.
(A) DMSO/RA dataset. The requirement for large absolute changes is illustrated by the solid blue lines. Consistency across conditions implied that genes must exhibit a change in the same direction in both treatments (bottom left or top right quadrant).
(B) R1–LR dataset. Note that the y-axis is the change seen at 72 h relative to that seen at 18 h. The requirement for large absolute changes is illustrated by the solid blue lines. The criterion for consistency was applied by requiring that the change 18 h after LIF removal be in the same direction as that after 72 h (i.e., in the lower left or upper right quadrants), regardless of its magnitude
(C) M–LR dataset. The requirement for large absolute changes is illustrated by the dashed blue lines. To meet the consistency criterion, we required that a temporal gene expression trend either increase or decrease continuously over the duration of the experiment. This requirement was relaxed slightly to retain trends with a direction change occurring either very early (red line) or very late (green line).