Pathway Activation

Documentation Index

Fetch the complete documentation index at: https://docs.revilico.bio/llms.txt

Use this file to discover all available pages before exploring further.

​

Why Use This Engine?

In the documentation below, we will use Revilico’s Pathway Activation assay engine to predict which oncogenic and stress-response signaling pathways are activated or suppressed by a drug across a cancer cell line panel. This assay provides a transcriptional-level view of the drug’s mechanism of action, identifying whether the compound engages DNA damage response, apoptotic signaling, survival pathways, or developmental programs, and at what concentrations these pathway changes occur.

​

Background

Cancer cells are driven by dysregulated signaling pathways that control proliferation, survival, and differentiation. When a drug perturbs these pathways, it triggers compensatory responses: survival pathways may be upregulated, apoptotic programs activated, and developmental pathways suppressed. Understanding which pathways are affected and in which direction helps predict both the therapeutic mechanism and the potential for adaptive resistance. RevAssay models eight core oncology signaling pathways drawn from KEGG and Reactome, covering the major kinase cascades relevant to solid tumor biology. The pathway database contains 47 pathways with curated KEGG IDs, Reactome IDs, gene targets, and confidence scores derived from literature concordance. Pathway fold-changes are derived from the cellular stress signal of the GNN+MLP viability model and calibrated to reflect the directional biology of stress-response signaling.

​

Simulation Model

For each pathway $p$, a direction (activated or inhibited) and a maximum fold-change scale factor $\alpha_p$ are defined: Activated pathways (fold-change above 1): $$\text{FC}_p = 1 + s \cdot (\alpha_p - 1) \cdot \eta_p, \quad \eta_p \sim \mathcal{LN}(0, 0.25^2)$$ Inhibited pathways (fold-change below 1): $$\text{FC}_p = 1 - s \cdot (1 - \alpha_p) \cdot \eta_p, \quad \eta_p \sim \mathcal{LN}(0, 0.20^2)$$ Where $s = 1 - v(c)$ is the cellular stress and $\alpha_p$ is the pathway-specific saturation fold-change. A dimensionless pathway activity score normalized to the range [0, 1] is derived from the fold-change: $$\text{score}_p = \min\left(1, \frac{|\log_2(\text{FC}_p)|}{3}\right)$$ A score of 0 indicates no change from baseline; a score of 1 corresponds to at least an 8-fold change in either direction. Modeled pathways and expected behavior under cytotoxic stress:

​

Parameters

​

Outputs

• Pathway fold-changes: Per-pathway FC at each drug concentration for each cell line
• Pathway activity scores: Normalized [0, 1] score per pathway
• Direction classification: Activated (red) or inhibited (blue) per pathway per concentration
• Horizontal bar chart: Pathways ranked by fold-change magnitude, colored by direction
• Dose-response curves: Per-pathway fold-change across the concentration range
• 96-well heatmap: Plate-view colored by selected pathway activity score
• Pathway metadata: KEGG ID, Reactome ID, gene targets, and confidence score for each modeled pathway