Drug safety.
Engineered.

Cytocast predicts potential clinical side effects from compound information and traces them to implicated off-targets and biological pathways, helping drug teams evaluate safety liabilities and investigate the mechanisms behind them.

Request the white paperExplore the platform

From compound to clinical insight

  1. Compound

    Structure-based input

  2. Off-target

    Binding predictions across 7,700+ proteins

  3. Perturbed pathway

    Propagation through protein complexes and pathways

  4. Tissue effect

    Simulated across 15 human tissues

  5. Predicted side effect

    1,200+ side-effect endpoints

One platform for every critical safety decision

Cytocast supports compound selection from early discovery to preclinical nomination, adapting the depth of analysis to each development stage.

For Pharma

Cytocast Screener

Identify early side-effect signals across large compound sets and eliminate weaker candidates before committing significant resources.

Cytocast Nominator

Build a decision-grade safety profile combining clinically relevant side-effect predictions with traceable biological mechanisms to support candidate advancement.

For BD & Investors

Cytocast Diligencer

Evaluate potential safety liabilities before you invest.

Cytocast Diligencer compares a drug’s predicted clinical side-effect profile with competing pipeline drugs and traces potential liabilities to implicated off-targets and pathways, supporting your investment or licensing evaluation.

HOW IT WORKS

Cytocast achieved higher ROC AUC than secondary pharmacology on 99 of 104 evaluated side-effect endpoints.

ROC AUC
ROC AUC: Secondary pharmacology and CytocastAxis: 0.5 to 1. Secondary pharmacology: 104 endpoints, median 0.658, quartiles 0.630 to 0.719, whiskers 0.601 to 0.851. Cytocast: 104 endpoints, median 0.775, quartiles 0.743 to 0.846, whiskers 0.614 to 0.976. All outliers are shown.0.50.60.70.80.91.00.658MedianSecondarypharmacology0.775MedianCytocast

Across 23 matched side-effect categories, Cytocast and animal toxicology had nearly identical median MCC scores.

MCC
MCC: Animal toxicology and CytocastAxis: 0 to 0.5. Animal toxicology: 23 endpoints, median 0.2958, quartiles 0.2429 to 0.3173, whiskers 0.1656 to 0.3834. Cytocast: 23 endpoints, median 0.2956, quartiles 0.2491 to 0.3492, whiskers 0.2090 to 0.4187. All outliers are shown.0.00.10.20.30.40.50.2958MedianAnimaltoxicology0.2956MedianCytocast
Methods and references

Boxes show the middle 50% of endpoint scores; white lines show medians; whiskers extend to observations within 1.5× IQR; dots show outliers. ROC AUC is the area under the receiver operating characteristic curve. MCC is the Matthews correlation coefficient.

The counts describe endpoints, not patients or prediction accuracy. Similar category-level medians do not establish statistical equivalence or replacement of animal testing.

  1. Secondary pharmacology: Sutherland, J.J., Yonchev, D., Fekete, A. et al. A preclinical secondary pharmacology resource illuminates target-adverse drug reaction associations of marketed drugs. Nature Communications 14, 4323 (2023).
  2. Animal toxicology: Steger-Hartmann, T., Raschke, M. Translating in vitro to in vivo and animal to human. Current Opinion in Toxicology 23–24, 6–10 (2020).