README.md

SingleCellExperiment

This package provides container class to represent single-cell experimental data as 2-dimensional matrices. In these matrices, the rows typically denote features or genomic regions of interest, while columns represent cells. In addition, a SingleCellExperiment (SCE) object may contain low-dimensionality embeddings, alternative experiments performed on same sample or set of cells. Follows Bioconductor's SingleCellExperiment.

Install

To get started, install the package from PyPI

pip install singlecellexperiment

Usage

The SingleCellExperiment extends RangeSummarizedExperiment and contains additional attributes:

• reduced_dims: Slot for low-dimensionality embeddings for each cell.
• alternative_experiments: Manages multi-modal experiments performed on the same sample or set of cells.
• row_pairs or column_pairs: Stores relationships between features or cells.

Readers are available to parse h5ad or AnnData objects to SCE:

import singlecellexperiment

sce = singlecellexperiment.read_h5ad("tests/data/adata.h5ad")

## output
class: SingleCellExperiment
dimensions: (20, 30)
assays(3): ['array', 'sparse', 'X']
row_data columns(5): ['var_cat', 'cat_ordered', 'int64', 'float64', 'uint8']
row_names(0):
column_data columns(5): ['obs_cat', 'cat_ordered', 'int64', 'float64', 'uint8']
column_names(0):
main_experiment_name:
reduced_dims(0): []
alternative_experiments(0): []
row_pairs(0): []
column_pairs(0): []
metadata(2): O_recarray nested

OR construct one from scratch

from singlecellexperiment import SingleCellExperiment

tse = SingleCellExperiment(
    assays={"counts": counts}, row_data=df_gr, col_data=col_data,
    reduced_dims={"tsne": ..., "umap": ...}, alternative_experiments={"atac": ...}
)

Since SingleCellExperiment extends RangeSummarizedExperiment, most methods especially slicing and accessors are applicable here. Checkout the documentation for more info.

Note

This project has been set up using PyScaffold 4.5. For details and usage information on PyScaffold see https://pyscaffold.org/.