Post 3 of 3
Introduction
This is the final article of a three-part series.
In the first article in this series, Decoding Chronic Pain Neurons, we introduced the recent Cell study that identified the molecular signature of “sleeping” nociceptors and highlighted Derek Howard’s role as co–first author.
In the second post, Making Sense of Rare Neurons Through Integration, Not Just Sequencing, we looked more closely at how that discovery was made, showing how integrating electrophysiology, spatial transcriptomics, cross-species datasets, and single-cell sequencing allowed researchers to connect neuronal behavior to molecular identity.
Together, those posts illustrate a broader shift underway in single-cell biology: moving beyond defining cells purely by what genes they express, and toward defining them by what they do, where they come from, or how they respond to change.
If you work in pharmaceutical R&D, you’re probably familiar with single-cell RNA-seq (scRNA-seq). And if you’re familiar with scRNA-seq, you’ve likely spent plenty of time staring at UMAP plots: clusters of cells colored by cell type, heatmaps showing marker genes that define each population.
But most biological questions are not really about what genes a cell happens to be expressing at one moment. They’re about mechanism. What controls a cell’s state? Why does one cell behave differently from another? Where did a cell come from, and where might it go next?
A growing class of single-cell methods addresses these questions directly by flipping the traditional experimental design. Instead of using gene expression to define a cell, these approaches define a cell by an intervention, a behavior, or an ancestry, and then use transcriptomics to interpret the result.
In this final post, we look at three examples of these function-first single-cell methods and what they reveal about the next phase of single-cell biology.
1. Perturb-seq: What Controls a Cell’s State?
Key papers: Dixit et al., Cell, 2016 | Adamson et al., Cell, 2016
The Setup
Perturb-seq combines pooled genetic perturbations (typically CRISPR knockouts or knockdowns) with single-cell RNA-seq. Each cell carries a known perturbation barcode, and the transcriptome reports the downstream consequences of that intervention.
In this design, the perturbation is the experiment. RNA-seq is the phenotype.
What It Tells You
Perturb-seq tells you about causality.
Traditional scRNA-seq asks: “Which genes are associated with this cell state?”
Perturb-seq asks: “If I disrupt gene X, does the cell move toward or away from this state?”
That distinction matters enormously. Association is not causation, and correlation structures in expression data can be misleading. With Perturb-seq, you can directly test:
- Which genes control a transcriptional program
- Whether different genes act in the same pathway
- Whether the same perturbation produces different outcomes in different cell types
Two cells with nearly identical baseline expression may respond very differently to the same perturbation. Perturb-seq exposes that hidden regulatory structure.
The Role of Transcriptomics
Here, RNA-seq isn’t discovering cell types. It’s measuring the effect of a controlled intervention. The readout is only interpretable because the perturbation is known.
2. Patch-seq: Why Does a Cell Behave the Way It Does?
Key papers: Cadwell et al., Nature Biotechnology, 2016 | Gouwens et al., Cell, 2020
The Setup
Patch-seq combines patch-clamp electrophysiology with single-cell RNA-seq from the same neuron. Electrical properties like firing rate, action-potential duration, adaptation dynamics are measured first. Gene expression is used to explain those functional differences.
This is one of the purest examples of a function-first design.
What It Tells You
Patch-seq tells you about functional identity.
Here’s why that matters: two neurons can cluster together transcriptionally, share all the same canonical marker genes, and yet behave completely differently.
Patch-seq reveals that they:
- Fire at different frequencies
- Adapt differently to repeated stimulation
- Have fundamentally different roles in circuit computation
And then transcriptomics reveals why! For example, through differential expression of sodium, potassium, or calcium channels.
The Role of Transcriptomics
RNA-seq provides a mechanistic explanation. It translates observed electrical behavior into molecular terms.
This is especially critical in systems like the brain, where function cannot be reliably inferred from markers alone. A cell’s job in a circuit depends on biophysics, and biophysics depends on precise channel composition. Patch-seq connects those dots.
3. Lineage Tracing: Where Did This Cell Come From?
Key papers: Wagner et al., Cell, 2018 | Weinreb et al., Science, 2020
The Setup
Lineage-resolved single-cell methods introduce heritable barcodes or record mutational histories that define ancestry relationships between cells. scRNA-seq is then used to profile their current transcriptional state.
Here, cellular history is the primary signal.
What It Actually Tells You
Lineage tracing asks whether transcriptionally similar cells share a past.
This lets you answer questions like:
- Do transcriptionally similar cells arise from the same progenitors, or do they converge from different lineages?
- Are fate decisions deterministic or flexible?
- Can cells with different origins reach the same endpoint?
This matters because two cells that look identical right now may have arrived there through completely different developmental paths and may have different future potential as a result.
The Role of Transcriptomics
Transcriptomics captures the current state. Lineage captures trajectory and constraint. Together, they separate appearance from origin.
The Shared Lesson
Despite their very different experimental designs, these three approaches share a core principle:
Gene expression is not the defining measurement, it’s the interpretive layer.
- Perturb-seq defines a cell by what was done to it
- Patch-seq defines a cell by how it behaves
- Lineage tracing defines a cell by where it came from
In all three cases, scRNA-seq is essential but it’s not sufficient on its own.
What This Means for How We Analyze Single-Cell Data
Across this series, a common theme has emerged. The most interesting discoveries are no longer coming from simply cataloging cell types. They are coming from experiments designed to explain function.
In the Cell study we discussed earlier in this series, researchers were able to identify sleeping nociceptors not because they sequenced more cells, but because they connected multiple measurements of the same biological system: electrophysiology, transcriptomics, spatial context, and cross-species datasets. The insight came from the integration.
Function-first approaches like Perturb-seq, Patch-seq, and lineage-resolved single-cell methods push this idea even further. Each introduces a new primary signal: a genetic intervention, a physiological behavior, or a cellular ancestry. Transcriptomics then becomes the layer that helps explain those signals mechanistically.
For bioinformatics, this creates a different class of analytical problems.
The challenge is no longer just clustering cells or identifying marker genes. Instead, it becomes about:
- Aligning fundamentally different data types (transcriptomes + electrophysiology, transcriptomes + perturbation phenotypes, transcriptomes + lineage histories)
- Preserving continuous biological signals rather than forcing everything into discrete clusters
- Designing models that can connect mechanism, behavior, and identity
These problems cannot be solved by standard single-cell pipelines alone. They require analytical strategies designed around the biology of the experiment, not just the structure of the sequencing data.
At Bridge Informatics, this is the kind of problem we work on every day. Our team helps pharmaceutical and biotech partners translate complex, multi-modal datasets into clear biological insight, whether that means integrating perturbation screens, spatial data, electrophysiology, or cross-species atlases.
Single-cell transcriptomics has given researchers an unprecedented view of cellular diversity. But diversity is only the starting point. The most important questions in biology are about causality, mechanism, and dynamics, and answering them increasingly requires integrating many different types of data into a coherent framework.
Conclusion
As single-cell technologies continue to evolve, studies will increasingly move beyond describing cells toward explaining how they function, change, and interact within complex biological systems. The methods we discussed here are early examples of that shift, and they highlight the analytical challenges that come with it. If you are ready to transform complex biological data into meaningful insight, click here to schedule a free introductory call with a member of our team.