Start with the data
- Data intake
- Quality control
- Filtering
scRNA-agent / By Lab Agent Works
Single-cell RNA-seq analysis, directed by you.
scRNA-agent helps biologists analyze their own data with local R/Python tools and specialized AI-assisted guidance, from quality control and annotation to figures and reports.
Built for biologists
Less distance between a question and an analysis.
Routine analysis can demand substantial computational expertise. scRNA-agent helps bridge that gap.
Interact with an agent that uses project context, executes local R/Python tools, and tracks outputs as your analysis develops.
It supports iterative biological exploration without replacing scientific judgment. Complex study design, unusual datasets, and challenging interpretation may still require expert bioinformatics.
The analysis framework
Guidance from data intake to interpretation.
Assess & prepare
- Doublet detection
- Ambient RNA handling
- Dimensionality reduction
Resolve the biology
- Batch / integration assessment
- Clustering
- Annotation
Explore & communicate
- Differential analysis
- Pathway analysis
- Figures and reports
An adaptable framework, not a fixed pipeline. The steps and methods depend on the dataset and biological question.
The system concept
A local workspace. A connected Core.
Local tools perform the computation. Your own model access supports reasoning and execution. The protected Core provides maintained scientific guidance and product services.
On your computerLOCAL WORKSPACE
01
Your scientific data
Raw matrices & processed inputs
02
scRNA-agent workspace
Your question, context & decisions
03
R / Python analysis
Local scientific computation
04
Project outputs
Figures, tables, reports & provenance
Raw matrices stay on your computer. Scientific computation runs locally.
YOUR MODEL ACCESS
Your own AI provider supports reasoning and execution. Prompts and shared project context may leave your computer and are subject to that provider’s policies.
An example interaction
Ask in the language of your research.
Researcher
“Compare the annotated T cells and myeloid cells using Hallmark GSEA and generate the plots.”
scRNA-agent / Illustrative workflow
- Prepare the comparison and clarify the analysis design.
- Acquire supported reference resources when needed.
- Run the analysis in the local scientific environment.
- Generate figures and reports for researcher review.
- Preserve project provenance and analysis artifacts.
An example of the intended interaction, not a scientific result. Method choices require review in the context of the study.
What you keep
Outputs that stay with your project.
Figures & tables
Visual summaries and tabular results for review and further exploration.
figures/ · tables/Analysis & provenance
Analysis artifacts, parameters, and records that keep the work traceable.
artifacts/ · provenance/Summaries & reports
Scientist-facing summaries alongside technical reports for deeper inspection.
summaries/ · reports/Illustrative output categories. Contents depend on the analysis performed.
Product capabilities
Made for careful, iterative work.
AutoRun & Guided Run
Generate an initial annotated dataset automatically, or work through analysis with researcher oversight.
Local R / Python execution
Run scientific tools in local environments, close to your data.
Interactive exploration with Codex
Ask new biological questions, revise earlier decisions and continue analyzing your annotated dataset.
Differential expression & enrichment
Investigate biological comparisons with differential expression, pathway enrichment and GSEA.
Reproducible figures and reports
Keep generated figures and reports connected to the analysis that produced them.
Reusable annotated datasets
Keep annotated datasets, figures and reports available for further investigation.
Protected scientific skills
Access specialized analysis skills through the protected Core service.
Traceable project outputs
Retain analysis artifacts, parameters, and provenance in your project.
Local-first data handling
Keep raw matrices and project files under your control.
Maintained workflow improvements
Benefit from developer-maintained guidance and reusable analysis experience.
Data handling
Keep raw scientific data local.
The Core does not need your raw dataset. Your model provider is a separate connection, with its own policies for shared context.
Read the privacy modelCommon questions
Before you get started.
Who is scRNA-agent for?
Primarily biologists who want to work directly with their single-cell RNA-seq data, with structured computational guidance. Complex study design and interpretation may still benefit from expert bioinformatics support.
Does it replace scientific judgment?
No. Researchers review analysis choices, assess quality, and interpret findings. AI assistance and workflow guidance do not guarantee scientific correctness.
Where does my data go?
Raw matrices and large scientific datasets stay in your local workspace. The protected Core supplies guidance and product services without needing the raw dataset. Your model provider is a separate service; review what context you share and its data policies.
Do I need my own AI access?
Yes. Connect and authenticate your own Codex access. Provider access requirements are separate from the free research beta.
What happens when the trial ends?
Existing project files and results are preserved. After the seven-day trial expires, new Workbench access requires authorization.
Can I download it now?
Not yet. The unsigned Windows candidate is awaiting final interactive Windows acceptance. The download page explains setup and the free seven-day research beta.
Your next biological question
Closer to your data.
Further in your research.
Meet scRNA-agent, the first tool from Lab Agent Works.