This repository contains the code for a modified version of scib package.
For original instruction, refer to the original repo and manuscript.
conda create -n leiden python=3.9
conda activate leiden
conda install cudatoolkit=11.7 -c pytorch -c nvidia
pip install torch==1.13.1+cu117 torchvision==0.14.1+cu117 torchaudio==0.13.1 --extra-index-url https://download.pytorch.org/whl/cu117
pip install --extra-index-url https://pypi.nvidia.com cudf-cu11==23.4.1 dask-cudf-cu11==23.4.1 cuml-cu11==23.4.1 cugraph-cu11==23.4.1 cucim==23.4.1
pip install einops ipdb pydance torchmetrics rapids-singlecell==0.9.6 scvi-tools wandb hdf5plugin
conda install ipykernel
git clone [email protected]:wehos/scib_gpu.git
cd scib_gpu
pip install -e .
Prepare an h5ad file, with embeddings stored in .obsm['X_emb'] or .X, batch labels in batch_label, cell type labels in cell_type.
Codes for standard evaluation.
INPUT_ANNDATA_PATH = 'xxx.h5ad'
OUTPUT_ANNDATA_PATH = 'yyy.csv'
import anndata as ad
import scib
import scanpy as sc
adata = ad.read_h5ad(INPUT_ANNDATA_PATH)
if 'X_emb' not in adata.obsm:
adata.obsm['X_emb'] = adata.X # If embeddings are stored in .X
if 'batch_label' not in adata.obs:
adata.obs['batch_label'] = adata.obs['batch']
### To specify resolution choices for leiden, call `cluster_optimal_resolution` manually before scib.metrics
if False: # Change to True to enable this code block
sc.pp.neighbors(adata, use_rep='X_emb')
scib.me.cluster_optimal_resolution(adata, cluster_key="cluster", resolutions=[0.1, 0.25, 0.5, 0.75, 1.0], label_key="cell_type")
###
res = scib.metrics.metrics(adata, adata, "batch_label", "cell_type", embed="X_emb", cluster_key="cluster", organism='human',
ari_ = True, nmi_ = True, silhouette_=True, isolated_labels_asw_ =True, isolated_labels_=True, pcr_=True,
graph_conn_ =True, lisi_graph_=True)
res.to_csv(OUTPUT_ANNDATA_PATH)
Luecken, M.D., Büttner, M., Chaichoompu, K. et al. Benchmarking atlas-level data integration in single-cell genomics. Nat Methods 19, 41–50 (2022). https://doi.org/10.1038/s41592-021-01336-8
The scib python package is available on PyPI and can be installed through
pip install scib
Import scib in python:
import scibThe package contains optional dependencies that need to be installed manually if needed.
These include R dependencies (rpy2, anndata2ri) which require an installation of R integration method packages.
All optional dependencies are listed under setup.cfg under [options.extras_require] and can be installed through pip.
e.g. for installing rpy2 and bbknn dependencies:
pip install 'scib[rpy2,bbknn]'
Optional dependencies outside of python need to be installed separately. For instance, in order to run kBET, install it via the following command in R:
install.packages('remotes')
remotes::install_github('theislab/kBET')Install pre-commit to the repository for running it automatically every time you commit in git.
pre-commit install
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent