Skip to main content
. 2017 Jan 25;4(1):73–83.e10. doi: 10.1016/j.cels.2016.11.013

Figure 4.

Figure 4

Non-canonical and Context-Specific Signaling

(A) Causal descendancy matrix showing causal effects observed under mTOR inhibitor AZD8055 in each of the 32 (cell line, stimulus) contexts. Rows represent phosphoproteins and columns represent contexts (see Figure 3). Black boxes indicate phosphoproteins that show a salient change under mTOR inhibition in a given context (see STAR Methods) and can therefore be regarded as causal descendants of mTOR in the signaling network for that context. The final column on the right indicates phosphoproteins that are descendants of mTOR in the canonical mTOR signaling pathway shown in (B). Phosphoproteins are ordered first by canonical column and then by hierarchical clustering of all data. See Figure S3 for causal descendancy matrices for the other inhibitor regimes.

(B) Canonical mTOR signaling pathway. Blue nodes are descendants of mTOR in the network, and white nodes are non-descendants. The pathway shown is a subnetwork of the prior network used within the network inference procedure (Figure S4). Full nodes names, including phosphorylation sites, are provided in Table S4.

(C) Summary of western blot validations of causal effects observed in RPPA data. A number of observations from the causal descendancy matrices were chosen for validation via western blot analysis. The number of phosphoprotein validations attempted (“Tested”) and the number of these that successfully validated (“Validated”) are presented for various (cell line, stimulus, inhibitor) combinations. Summary totals are also presented for each cell line, each inhibitor, and across all validation experiments. See also Table S2.