Accurate and fast cell marker gene identification with COSG¶

Overview¶

COSG is a cosine similarity-based method for more accurate and scalable marker gene identification.

COSG is a general method for cell marker gene identification across different data modalities, e.g., scRNA-seq, scATAC-seq and spatially resolved transcriptome data.
Marker genes or genomic regions identified by COSG are more indicative and with greater cell-type specificity.
COSG is ultrafast for large-scale datasets, and is capable of identifying marker genes for one million cells in less than two minutes.

The method and benchmarking results are described in Dai et al., (2021).

Documentation¶

The documentation for COSG is available here.

Tutorial¶

The COSG tutorial provides a quick-start guide for using COSG and demonstrates the superior performance of COSG as compared with other methods, and the Jupyter notebook is also available.

Question¶

For questions about the code and tutorial, please contact Min Dai, daimin@zju.edu.cn.

Citation¶

If COSG is useful for your research, please consider citing Dai et al., (2021).

Usage¶

Import COSG as:

import cosg as cosg

amd import Scanpy as:

import scanpy as sc

Next, load the data via:

adata = sc.datasets.pbmc68k_reduced()

then identify marker genes for each cell group by running:

cosg.cosg(adata, key_added='cosg', groupby='bulk_labels')

and the top marker genes can be visualized via:

sc.pl.rank_genes_groups(adata, key='cosg')

Installation¶

Runnig this package requires a Python environment (>=3.6).

Development Version¶

To use the latest version on GitHub: please clone the repository and cd into the root directory, and run:

pip install -e .

PyPI¶

Please run:

pip install cosg

API¶

Import cosg as:

import cosg as cosg

Marker gene identification¶

Release notes¶

Version 1.0¶

1.0.0¶

COSG is an accurate and efficient marker gene identification method for single cell sequencing data.

COSG is applicable to single-cell RNA sequencing data, single-cell ATAC sequencing data and spatially resolved transcriptome data.
COSG is fast and scalable for ultra-large datasets of million-scale cells (less than two minutes for one million cells with one CPU core).
Marker genes or genomic regions identified by COSG are more indicative and with greater cell-type specificity.

Min Dai (2021-06-15)