SCORED 0.1.0

A package for single-cell RNA sequencing data imputation using SimRank and Random Walk with Restart.

SCORED

SCORED is a Python package for single-cell RNA sequencing data imputation using SimRank and Random Walk with Restart algorithms. It processes AnnData objects, performs quality control, builds graphs, and imputes missing data.

Installation

pip install SCORED

Usage

import scanpy as sc
import torch
from scored import SCORED

# Load your AnnData object
adata = sc.read_h5ad("your_data.h5ad")

# Run SCORED
imputed_matrix = SCORED(
    adata_tr=adata,
    condition_key="condition",
    device="cuda" if torch.cuda.is_available() else "cpu"
)

Requirements

• Python >= 3.8
• scikit-learn >= 1.0
• scipy >= 1.7
• matplotlib >= 3.4
• seaborn >= 0.11
• tqdm >= 4.0
• networkx >= 2.6
• numpy >= 1.20
• torch >= 1.9
• scanpy >= 1.8
• pandas >= 1.3

License

MIT License