We combine AI and mechanistic modelling to build biological networks from your data and quantify cell-specific treatment effects. Biology is no average - we help you build explainable cell-state conditioned biological models.

We help to turn your single-cell (phospho)-protein data into actionable insight on how drug-treatments effect signalling networks and cell-fate — resolving signalling at the single-cell level. We model signalling pathways conditioned on the state of individual cells/ patients to better explain and predict outcome of treatment strategies. Our platform combines mechanistic modelling of pathways from (single-cell) protein data and existing knowledge of signalling pathways to train AI models to predict how various cell-states/ types or patients responde to treatment. Our platform does not treat signalling pathways as a black-box, but provides actionable insight into the signalling dynamics that govern treatment response.

Not a black box. We start from prior knowledge of signalling networks and refine it with data-driven edges — using our own single-cell network reconstruction methods to explain differences between cell states with interpretable, mechanistic structure.

One model across the full cell-state continuum. We train an AI model jointly across the entire continuum of cell states in your single-cell dataset to predict how cells in each individual state respond to drug treatment.

Human in the loop. Bring your own knowledge of interactions and their strengths into the model, then scan the single-cell space yourself with our software — inspecting how signalling changes across it and how drug treatment shifts signalling in the states you pick.

The full platform is on its way. Stay tuned.

info@signallingintelligence.com