Decode signalling
at single-cell resolution

Unravel and understand signalling from your protein data! We extract, map, and interpret intracellular signalling programs from single-cell protein data — turning it into actionable pathway biology for drug discovery. By combining AI-driven and mechanistic models, we deliver explainable insight into the signalling mechanisms that drive each cell's behaviour within your protein data. We are working with various protein measurement technologies to unravel signalling from bulk patient data or single-cell data.

Request a demo See the platform →

Differences in signalling drive cell fate — distinct clusters run distinct pathway programs. We recover those signalling differences for you, so you can understand drug effects at the signalling level and better predict potential targets by leveraging the flow of information through the network.

Cluster A — MAPK / ERK branch
Cluster B — PI3K / AKT branch
Crosstalk between branches
01 / DEMUX
Processing of raw sequncing data
We use our own demultiplexing platform (ESGI, see papers below), which handles any barcoded single-cell sequencing technology with enhanced read-recovery accuracy and detailed error reports for fine-tuning experimental design.
02 / NORM
Normalisation of single-cell count data
We use our own state-of-the-art normalization method. It mayorly improves normalization comapred to current normalization standards and separates biological concentration from technical abundance — correcting library size and cell-size variation simultaneously.
03 / CORR
Local correlation analysis
We use our own tool LoCo (Local Correlation) to find locally co-varying protein pairs within cell-state neighbourhoods, resolving relationships invisible to global correlation methods. These interpretable patterns are used in our own AI-model to make further predictions on cell-state specific signalling cahnges.
04 / NET
Pathway inference
Correlation programs are inferred per population with scCNR/scMRA (see papers below). This algorithm then gives us mechanistic insight into cell-state specific signalling networks. This insight is then further analyzed by our own AI platform to give further insight into the signalling networks on the single-cell level.
Why phosphoproteomics

Transcriptomics tells you what genes are on.
We tell you what the cell is doing.

Proteins are where biology actually happens — they are the enzymes, receptors, and machines that carry out a cell's decisions. Messenger RNA is only a blueprint, and it is a famously unreliable one: transcript levels correlate weakly with the amount of protein that is finally made, so an mRNA readout is a proxy for a proxy.

Phosphorylation goes one step further. It captures which proteins are switched on right now — the live activity state of a signalling network, minutes ahead of any change in gene expression. That makes single-cell proteomics a far more truthful readout of what a cell is doing than scRNA-seq can ever be.

A proxy vs. the actor
mRNA abundance predicts protein levels poorly. The phosphoproteome measures the functional molecules themselves, not a transcriptional guess at them.
Activity, not just presence
Phosphorylation reports kinase signalling state in real time — the difference between a protein being present and a pathway being active.
Closer to the phenotype
The signalling rewiring that drives drug response plays out at the protein level, often completely invisible to transcriptomics.
Further reading

Scientific papers

The SignallingIntelligence (SI) platform grew out of Tim Stohn's PhD thesis, which produced a suite of tools for analyzing single-cell phosphoprotein data. SI builds on this foundation by combining several peer-reviewed tools — which generate mechanistic input — with AI-driven models that refine our understanding of signalling, without sacrificing the interpretability of the mechanistic insight. Here you can find peer-reviewed methods and preprints behind our demultiplexing, normalisation, and signalling-network inference methods.

Bioinformatics
Our pathway inference method — Reconstructing and comparing signal transduction networks from single-cell protein quantification data
Stohn, van Eijl, Mulder, Wessels & Bosdriesz · Bioinformatics · 2026
→ Mol Cell Proteomics
Application example of our pathway inference method — Cell-State-Specific Drug Responses are Associated With Differences in Signaling Network Wiring
Krämer, van Eijl, Stohn, Tanis, Wessels3, Bosdriesz, Mulder · Molecular & Cellular Proteomics · 2026
→ Preprint · bioRxiv
Our generic sequencing processing tool — ESGI: Efficient splitting of generic indices in single-cell sequencing data
Stohn, van de Brug, Theodosiadou, Thijjssen, Jastrzebski, Wessels, Bosdriesz · bioRxiv · 2026
→
FAQ

Single-cell signalling, explained

Common questions about single-cell signalling analysis, phosphoprotein network inference, and how SignallingIntelligence turns single-cell proteomics into actionable pathway biology.

What is single-cell signalling analysis?
Single-cell signalling analysis measures and interprets intracellular signal transduction in individual cells. Instead of averaging across a whole population, SignallingIntelligence resolves how signalling pathways differ between cell states using single-cell protein and phosphoprotein data — showing what each cell is actually doing, not just which genes are transcribed.
How do you infer signalling networks from single-cell proteomics data?
We map co-varying protein and phosphoprotein measurements onto curated interaction databases such as SIGNOR and OmniPath to reconstruct active signal transduction and phosphoprotein networks. Our peer-reviewed pathway-inference method (Bioinformatics, 2026) reconstructs and compares these networks across cell states to reveal which signalling programs are active and how they are wired.
What are Phospho-seq and SIGNAL-seq?
Phospho-seq and SIGNAL-seq are single-cell technologies that jointly profile protein, phosphoprotein and transcriptomic readouts. Phosphorylation captures which proteins are switched on in real time — the live activity state of a signalling network — making single-cell phosphoproteomics a more direct readout of cell behaviour than scRNA-seq.
What do our tools do?
ESGI handles combinatorial barcode demultiplexing for single-cell (phospho)protein data, but can be applied to any sequencing based barcoded data like spatial data, etc. , LoCo (local correlation) finds co-varying protein pairs within cell-state neighbourhoods, resolving relationships invisible to global correlation. Together they turn raw single-cell data into interpretable signalling programs.
How does single-cell phosphoproteomics reveal drug effects and new targets?
Drug response and resistance are driven by rewiring of signalling networks at the protein level, which is often invisible to transcriptomics. By quantifying phosphoprotein network activity across cell states, we identify how a treatment changes signalling flow and pinpoint the nodes — potential therapeutic targets — that drive each cell's response.
Get in touch

Let's make sense of
your dataset

Want to decode signalling in your single-cell data? Reach out to see how our platform can help you disentangle signalling effects at the single-cell level.

Tim Stohn
Scientific Advisor
Tim Stohn
SignallingIntelligence · Leiden
t.stohn@signallingintelligence.com
Ezgi Arslantürkoğlu
Business Development & Partnerships
Ezgi Arslantürkoğlu
SignallingIntelligence · Leiden
e.arslanturkoglu@signallingintelligence.com