Measure the similarity of cells from 2 different single-cell RNAseq datasets.
scCompare provides cell similarity measurements between 'test' and 'mapping' scRNAseq datasets. It is agnostic of known cell types, taking into account only data from the 'mapping' dataset. It includes:
- plots to visualize similarity and correlation between cells
- 'unmapped' annotations for cells that do not pass a similarity cutoff
- statistical metrics
Some nomenclature that may be non-intutitve is described here:
user annotation = user-defined cell annotations in the mapping dataset
assigned annotation (asgd_cluster) = annotation with cutoff
raw annotation (canon_label_asgd) = annotation without cutoff
Make sure you are using a recent version of pip, so first
pip install -U pip
then
pip install git+https://github.com/bluerocktx/bfx-scCompare
- Check out our vignette here to walk through an example of how to use scCompare.