11. Robustness to realistic data problems

11. Robustness to realistic data problems

 figure
Figure 15. The 14 responses absent from IID+TRRUST — interactome incompleteness.
 figure
Figure 16. Edge provenance: 1,886,194 directed PPI edges plus 5,551 unique directed TRRUST regulatory edges (full network).

Two realistic data problems occur in the run itself: 14/182 responses are missing from IID+TRRUST, and directed TRRUST edges could be wrong. Dropping all 531 subnetwork TRRUST edges moves ranks by rho = 0.945 (top-20 16/20) — regulatory direction matters, but the ranking survives its removal. The TRRUST directions here are annotations, not random: this ablation is a genuine test of their contribution. Sign of regulation is still ignored throughout.

Assay supplement ()

Dissemination readout
Dissemination readout
Sema5a 0% (Gö-6976) to Col6a1 94.7%.
IC50 distribution
IC50 distribution
Nanomolar potencies from the dissemination screen.
Potency vs effect
Potency vs effect
IC50 against dissemination.
Assay table
Assay table
All 18 genes with a dissemination readout.
Assay panels (strongest)
Assay panels (strongest)
dissemination % with NetPert ranks, strongest 9 of 18.
Assay panels (weakest)
Assay panels (weakest)
dissemination % with NetPert ranks, weakest 9 of 18.
Assay waterfall
Assay waterfall
inhibition ordered with NetPert ranks annotated.