Learn single-cell data structure through topological bases, graphs and layouts
Project description
CellTOMetry: single-Cell Topologically Optimized Geometry
CellTOMetry is a python library to orchestrate topological single-cell data analysis. It is centered around TopOMetry and scanpy.
Installation and dependencies
CellTOMetry requires scanpy and TopOMetry. After installing both, install celltometry with:
pip3 install celltometry
Using CellTOMetry with scanpy
This is a quick-start. For further instructions, check TopOMetry documentation.
First, we load libraries and some data to work with:
import scanpy as sc
import topo as tp
import celltometry as ct
# Load the PBMC3k dataset
adata = sc.datasets.pbmc3k()
Next, we perform the default preprocessing workflow with scanpy: libraries are size-normalized, log-transformed for variance stabilization, and subset to highly variable genes.
# Normalize and find highly variable genes
adata = ct.preprocess(adata)
Then, we proceed to the default scanpy workflow. It corresponds to:
- Scaling data (optional, changes adata.X) - ``
- Performing PCA
- Learning a neighborhood graph
- Learn an UMAP projection with this graph
- Cluster this graph with the Leiden community detection algorithm
Similar to preprocessing, we wrap it with an one-liner:
adata = ct.default_workflow(adata, scale=True)
To run the topological workflow, create a TopOGraph object tg
and use it to learn and add information to AnnData
:
adata = ct.topological_workflow(adata, tg)
For further instructions, please check TopOMetry documentation.
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
Built Distribution
Hashes for celltometry-0.0.1-py3-none-any.whl
Algorithm | Hash digest | |
---|---|---|
SHA256 | 70d39af8a7ab5fbf2797e00a35f15d497214464197e7d9ff8b61c710263bb732 |
|
MD5 | 2d70d0fa0043c7ee893385846610d3dd |
|
BLAKE2b-256 | bc35e1cf300f0984b3e8991ddb1ac7bed419f561f8bf375b8c05a404d1c4c4ad |