README.md

scquill

Approximate any single cell data set, saving >99% of memory and runtime.

It's pronounced /ˈskwɪɹl̩//, like the animal.

Approximating a single cell data set

import scquill

q = scquill.Compressor(
    filename='myscdata.h5ad',
    output_filename='myapprox.h5',
    celltype_column="cell_annotation",
)

q()

Exploring an approximation

To load an approximation:

import scquill

app = scquill.Approximation(
    filename='myapprox.h5',
)

To show a dot plot:

scquill.pl.dotplot(app, ['gene1', 'gene2', 'gene3'])

To show a neighborhood plot:

scquill.pl.neighborhoodplot(app, ['gene1', 'gene2', 'gene3'])

To show embeddings of cell neighborhoods, similar to single-cell UMAPs:

scquill.pl.embedding(app, ['gene1', 'gene2', 'gene3'])

MORE TO COME

Authors

Fabio Zanini @fabilab