WHAT SAME-CELL DATA REVEALS
Move from correlation toward biological causality
A cell's molecular state at any moment is the product of its trajectory: the signals it received, the perturbations it survived, the regulatory decisions that came before. Read that state once and you have an identity, not a trajectory.
Fate is decided in the individual cell, so it can only be decoded there. A population average suppresses the variation between nominally identical cells as noise. In a same-cell record that variation is the signal: the difference between cells that resist, adapt or differentiate and cells that do not.
Live-seq connects molecular state, perturbation, phenotype and outcome in one longitudinal record of the same cell. It does not replace bulk sequencing, scRNA-seq, imaging or perturbation screens. It anchors them.
Causality
See what causes change, not just what correlates with it, because the same cell is measured before and after the intervention.
Prediction
Link a cell’s initial state to its future fate: which cells resist, adapt, or differentiate, and what in their baseline decided it.
Mechanism
Follow pathways as dynamic processes in one cell, turning black-box correlations into interpretable biology.
Phenomics
Match the transcriptome with live-cell imaging of the same cell: molecular state bound directly to morphology and behavior.
Ground truth
Anchor points that unify bulk, scRNA-seq, imaging, and model predictions, because the trajectory is measured, not inferred.
RESEARCH APPLICATIONS
Where Live-seq changes the answer
Live-seq is most valuable when the biological question depends on what happens to the same cell over time. These are the places where that changes the answer.
LONGITUDINAL WORKFLOW
The measured trajectory of one living cell.
Sequencing does not have to disrupt a cell's fate.
A longitudinal Live-seq experiment follows the same basic logic: image the living cell, establish a baseline, introduce a perturbation, then return to the same cell to measure what changed.
PHENOTYPE ON RECORD
1. Image the living cell
The cell is selected under the microscope, so its morphology, position, markers and behavior are recorded before anything is taken from it.
That record is what every later readout attaches to.
BASELINE TRANSCRIPTOME
2. Read the baseline
A FluidFM Nanosyringe pierces the membrane and withdraws a small sample of cytoplasm that also contains RNA.
Sequencing it gives a genome-wide baseline.
THE INTERVENTION
3. Apply the perturbation
The same cell can be biopsied before and after a perturbation: a drug, a signal, a stress, a pathway modulator, or another controlled stimulus.
The intervention is applied to the cell you have already baselined.
THE SECOND READ
4. Read the same cell again
The same cell is read again, either by imaging it to its outcome or by taking a second biopsy.
You see the molecular trajectory and the fate outcome of that cell. Not an average population-level response.
LIVE-SEQ EXPERIMENTAL MODES
Three questions. Three scientific modes. Each mode answers a different causal question about your cells.
| THE QUESTION | HOW IT WORKS | WHAT YOU GET |
EXPLAIN “What makes this cell different?”
| See a phenotype. Biopsy that cell. Read the molecular signature behind it. |
The cell is selected through the microscope, so the phenotype is on record before the sample is taken and the transcriptome attaches to an observed behavior. Differential gene expression of a chosen phenotype, one transcriptome per cell you picked out. |
PREDICT “What will this cell become?” | Measure a cell's molecular state today. Watch its fate tomorrow. Find the predictors. |
Differential gene expression of future phenotypical outcomes: cells that looked identical at baseline, grouped retrospectively by what they became, with the molecular difference on record before anything visible appears. |
TEMPORAL “How does the story unfold?” | Two biopsies, one cell, one perturbation: the molecular state before and after. |
Both reads come from the same cell, so what you get is longitudinal differential gene expression, a change measured inside one individual. Single cell heterogeneity becomes the thing you are measuring. The trajectory is measured, not inferred. |
LIVE-SEQ EXPERIMENTAL MODES
Three questions. Three scientific modes.
Each mode answers a different causal question about your cells. Hover to learn more.
EXPLAIN: “WHAT MAKES THIS CELL DIFFERENT?"
See a phenotype. Biopsy that cell. Read the molecular signature behind it.
The cell is selected through the microscope, so the phenotype is on record before the sample is taken and the transcriptome attaches to an observed behavior.
What you get is differential gene expression of a chosen phenotype, one transcriptome per cell you picked out.
PREDICT: “WHAT WILL THIS CELL BECOME?”
Measure a cell's molecular state today. Watch its fate tomorrow. Find the predictors.
What you get is differential gene expression of future phenotypical outcomes: cells that looked identical at baseline, grouped retrospectively by what they became, with the molecular difference on record before anything visible appears.
TEMPORAL: “HOW DOES THE STORY UNFOLD?”
Two biopsies, one cell, one perturbation: the molecular state before and after.
Both reads come from the same cell, so what you get is longitudinal differential gene expression, a change measured inside one individual.
Single cell heterogeneity becomes the thing you are measuring. The trajectory is measured, not inferred.
METHOD VALIDATION
The cytoplasmic biopsy leaves the cell's biology intact, and reads its transcriptome genome-wide
Live-seq was developed by the Vorholt group at ETH Zurich and the Deplancke lab at EPFL, published in Nature in 2022, building on FluidFM intracellular sampling pioneered at ETH Zurich.
It rests on 2016 work showing a cytoplasmic biopsy can be physically retrieved from a cell that remains viable, and supplies the readout: a genome-wide transcriptome from that same sample, with the cell still available afterward.
THE BIOPSY ·
2016 · CELL
Tunable Extraction
FluidFM Nanosyringe aspirated a controlled cytoplasm volume. Cells stayed alive five days later and divided on schedule.
THE CONTENT · 2017 · ANALYTICAL CHEMISTRY
Metabolites, by mass spectrometry
Metabolites withdrawn from living cells by FluidFM were analyzed by mass spectrometry, providing opportunities for complementary analyses of the cell before, during, and after analysis.
THE SEQUENCING ·
2022 · NATURE
Live-cell sequencing
An ultra-low input RNA-seq protocol turned a few-picogram biopsy into a genome-wide transcriptome, cell still alive and sampleable again.
WHY LIVE-SEQ IS DIFFERENT
Not another single-cell snapshot
Conventional single-cell methods provide powerful population-level measurements, but destructive workflows lose the biological history of each individual cell. Live-seq preserves that continuity.
| Capability | scRNA-seq / Perturb-seq | Spatial Transcriptomics |
Live-seq |
| Cell status after measurement | Dead (lysed) | Dead (fixed) |
Alive |
| Longitudinal measurement | Impossible (cell destroyed) | Impossible (tissue fixed) |
Yes — same cell, multiple time points |
| True causal trajectories | No — pseudotime inference | No — spatial snapshots |
Yes — measured directly |
| Throughput per run | 10,000+ cells | 1,000+ spots |
~40 biopsies/day |
| Information per cell | One snapshot | One snapshot + spatial context |
Temporal trajectory + phenotype |
| Paired with intervention | Perturb-seq: population average | No |
Yes — single-cell precision |
| Multi-modal | Molecular only | Morphology and molecular |
Full: Molecular Phenotypical and functional (e.g., behavior) |
AI training value | Level 1 (correlations) | Level 1 (correlations) | Level 3 (counterfactual causal data) |
DOWNSTREAM ANALYSIS
Fits into your existing sequencing workflow
Live-Seq output is standard Illumina-compatible single-cell RNA sequencing data. Library preparation is powered by LuthOR, Lexogen’s ultra-low RNA input protocol designed for biopsy-based sampling.
