This project aims to provide an automatized pipeline for gating cell populations in fcs files.
Updates
- Fixed bug with subseting populations from parent populations by index
- Modified logical transformation for clearer use
- CSV report on identified cells frequencies added for further analysis
- Loop added to perform whole fcs folder analysis
Challenges
- Logicle transformation in flowCore differs from the flowJo version
- The smaller population the harder it is to gate it
- Unable to sort files with 1e7 cells and more (not enough cache)
- Algorithm fails to stably identify populations in random fcs file
This project would allow performing gatings of fcs files in automized fashion. Currently all gatings are done manually, which can prove to be time consuming task.
Pipeline was inspired by this paper that has introduced the flowDensity algorithm that
facilitates reproducible, high-throughput analysis of flow cytometry data by automating a predefined manual gating approach
It performed best in flowCAP-III competition.
Algorithm is based on estimation of regions surrounding cell populations via using such characteristics of the marker density distribution as number, height, width of peaks and the slope of a distribution curve.
 |
|---|
| flowDensity perfomance sneak peak |
AutoPipeline.R provides full gating of 1 given fcs file based on this gating strategy. Auto_pipeline_loop.R provides same gating but for a folder of given fcs files.
 |
|---|
| Example of good gating of Macrophages using flowDensity and lab's FCSs |
 |
|---|
| Example of bad gating of CD8, CD4 using flowDensity and lab's FCSs |
## Contributing
Pull requests are welcome. For major changes, please open an issue first
to discuss what you would like to change.
Please make sure to update tests as appropriate.