Cytokine Release

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 Cytokine Release assay engine to predict the inflammatory cytokine response to drug treatment across a cancer cell line panel. This assay identifies drugs at concentrations that induce excessive cytokine secretion, flagging cytokine storm risk before experimental testing. Cytokine storm is a potentially fatal immune cascade associated with high-dose cytotoxic therapies and certain immunomodulatory agents.

​

Background

Cytokines are small signaling proteins secreted by cells in response to stress, damage, or immune activation. When a drug induces significant cellular stress or direct immune activation, cytokine levels in the cell culture supernatant rise in proportion to the magnitude of the stress response. At very high drug concentrations or in immunocompetent contexts, cytokine levels can reach thresholds associated with systemic inflammatory toxicity. The Cytokine Release module models six core cytokines: IL-6, IL-8, IL-1 beta, TNF-alpha, IFN-gamma, and IL-10. The database encompasses over 60 catalogued cytokines spanning the interleukin, chemokine, interferon, TNF superfamily, colony-stimulating factor, and growth factor families. Cytokine levels are derived from the cellular stress signal generated by the underlying GNN+MLP viability model.

​

Simulation Model

For each cytokine, a baseline secretion level at zero drug effect and a maximum fold-induction at saturating stress are defined from literature-curated values for typical cancer cell lines: The cytokine level at stress $s = 1 - v(c)$ is computed as: $$[\text{Cytokine}](c) = B_k \cdot \left(1 + s \cdot (F_k - 1)\right) \cdot \eta, \quad \eta \sim \mathcal{LN}(0, 0.18^2)$$ Where $B_k$ is the baseline for cytokine $k$, $F_k$ is the maximum fold-induction, and $\eta$ is a lognormal multiplicative noise term with 18% coefficient of variation reflecting inter-well variability. Cytokine Storm Flag A cytokine storm event is flagged when any cytokine exceeds a per-cytokine threshold (default 500 pg/mL): $$\text{Storm} = \bigvee_k \left[ [\text{Cytokine}_k](c) > \theta_k \right]$$ The storm flag is reported per concentration per cell line, enabling identification of the minimum storm-inducing concentration. At clinical exposure concentrations (typically 1 to 5 micromolar for most kinase inhibitors), storm risk is generally absent. Storm flags at extreme concentrations (100 micromolar) reflect the acute toxicity profile rather than the therapeutic window.

​

Parameters

​

Outputs

• Cytokine levels (pg/mL): Per-cytokine concentration at each drug dose for each cell line
• Fold-induction: Ratio of drug-treated to untreated secretion per cytokine
• Cytokine storm flag: Binary indicator per concentration per cell line with minimum storm dose
• Multi-cytokine heatmap: All cytokines by concentration, colored by level relative to threshold
• Dose-response curves: Per-cytokine secretion curves across the dose range
• 96-well heatmap: Plate-view colored by selected cytokine level or storm risk
• Cytokine family breakdown: Grouped view across interleukins, chemokines, interferons, and other families