Methods used in Metaflow
Metaflow strives to be transparent on what is happening behind the scenes. We list below, the list of methods and, if applicable, their DOI used in Metaflow. Consider citing them in your work if you use them in your works!
- Pre-processing .fcs files:
- Transformation: Adjusts the scale of data for ease of downstream analyis. Users have the option to choose between the following transformation methods.
- Arcsinh (generalized hyperbolic sine)
- Biexponential
- Logicle (biexponential with dynamic parameter tuning) [DOI]
- MetaClean [DOI: coming soon!]: Removes low quality events in data that occur when there is clogging, bubbles, or other obstruction to flow while user samples were being processed in the cytometer.
- Transformation: Adjusts the scale of data for ease of downstream analyis. Users have the option to choose between the following transformation methods.
- Clustering [DOI: coming soon!]: Clusters cells into their respective clusters or groups based on the given high-dimensional data; also creates a clustering hierarchy that the user can explore by adjust the depth.
- Meta-analysis:
- Batch effect removal using PytoNorm, the Python version of CytoNorm [DOI]: Removes batch effects across files if the user selects for it to be done. This does NOT affect downstream statistics calculations.