Skip to main content
. 2015 Mar 3;11(3):0790. doi: 10.15252/msb.20145644

Figure 2.

Figure 2

Bayesian network dependency models infer connections between shape and transcription factor localization using intrinsic heterogeneity of single cells
  1. Left: Distribution of nuclear/cytoplasmic NF-κB ratios in single cells ± TNFα (blue = unstimulated, red = TNFα 1 h, green = TNFα 5 h) showing distribution of values for each cell line. Right: Distribution of first principal component (PC) of morphology features showing multi-modal distribution of cell shapes.
  2. Example Bayesian network (MCF7). Edge (line/arrow) color denotes treatment conditions. Arrows indicate direction of dependency, and lines indicate interactions where direction cannot be determined. Numbers denote confidence (see Materials and Methods).
  3. Frequency of NF-κB ratio dependencies by cell line for the most commonly connected features ± TNFα (1 h).
  4. NF-κB dependencies on morphological factors detected in each condition. Red = dependency detected in unstimulated condition. Asterisk = dependency detected in TNFα-stimulated cell (1 h).