SUREv1 1.0.0

Succinct Representation of Single Cells

SURE: SUccinct REpresentation of cells

SURE introduces a vector quantization-based probabilistic generative model for calling metacells and use them as landmarks that form a coordinate system for cell ID. Analyzing single-cell omics data in a manner analogous to reference genome-based genomic analysis.

$$\color{red}\text{\textbf{UPDATE}}$$

An update has been distributed. Users can access to it via SUREv2. It provides Python classes that users can call SURE in scripts. It also provide the command that users can run SURE in the shell. Additionally, SUREv2 supports the calling of metacells for multi-omics datasets.

Installation

1. Create a virtual environment

conda create -n SUREv1 python=3.10 scipy numpy pandas scikit-learn && conda activate SUREv1

2. Install PyTorch following the official instruction.

pip3 install torch torchvision --index-url https://download.pytorch.org/whl/cu126

3. Install SURE

pip3 install SUREv1

Example 1: Calling metacells for a single-cell dataset

Users can refer to here for details.

Example 2: The hierarchical assembly of large-scale dataset(s)

Users can refer to here for details.

Example 3: Human brain cell atlas

Users can refer to here for details.

Example 4: Metacell calling for scATAC-seq data

Users can refer to here for details.