integration
Project description
scbean.VIPCCA
Variational inference of probabilistic canonical correlation analysis
introduction......
............
Create conda enviroment
$ conda create -n scbean python=3.6
$ conda activate scbean
For more information, see https://docs.conda.io/projects/conda/en/latest/user-guide/concepts/environments.html
Install VIPCCA from pypi
$ pip install scbean
Install VIPCCA from GitHub source code
$ git clone https://github.com/jhu99/scbean.git
$ cd ./scbean/
$ pip install .
Note: Please make sure that the pip
is for python>=3.6. The current release depends on tensorflow with version 1.15.4. Install tenserfolow-gpu if gpu is avialable on the machine.
Usage
https://vipcca.readthedocs.io/en/latest/
Quick Start
Download example data at http://141.211.10.196/result/test/papers/vipcca/data.tar.gz
import scbean.vipcca as vip
from scbean import preprocessing as pp
from scbean import plotting as pl
# read single-cell data.
adata_b1 = pp.read_sc_data("./data/mixed_cell_lines/293t.h5ad", batch_name="293t")
adata_b2 = pp.read_sc_data("./data/mixed_cell_lines/jurkat.h5ad", batch_name="jurkat")
adata_b3 = pp.read_sc_data("./data/mixed_cell_lines/mixed.h5ad", batch_name="mixed")
# pp.preprocessing include filteration, log-TPM normalization, selection of highly variable genes.
adata_all= pp.preprocessing([adata_b1, adata_b2, adata_b3])
# VIPCCA will train the neural network on the provided datasets.
handle = vp.VIPCCA(
adata_all,
res_path='./results/CVAE_5/',
split_by="_batch",
epochs=100,
lambda_regulizer=5,
)
# transform user's single-cell data into shared low-dimensional space and recover gene expression.
adata_integrate=handle.fit_transform()
# Visualization
pl.run_embedding(adata_integrate, path='./results/CVAE_5/',method="umap")
pl.plotEmbedding(adata_integrate, path='./results/CVAE_5/', method='umap', group_by="_batch",legend_loc="right margin")
pl.plotEmbedding(adata_integrate, path='./results/CVAE_5/', method='umap', group_by="celltype",legend_loc="on data")
Project details
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
scbean-0.2.6.tar.gz
(1.9 MB
view hashes)
Built Distribution
scbean-0.2.6-py3-none-any.whl
(17.6 kB
view hashes)