Skip to content

README.md

Latest commit

 

History

History
136 lines (101 loc) · 3.67 KB

README.md

File metadata and controls

136 lines (101 loc) · 3.67 KB

PHLOWERlogo

decomPosition of the Hodge Laplacian for inferring trajectOries from floWs of cEll diffeRentiation

Hardware requirements

PHLOWER module requires a standard computer with enough RAM to support the in-memory operations, high performance computing (HPC) cluster is recommended.

System Requirements

PHLOWER has been tested with the following OS or virtual environment:

- macOS Sonoma 14.5
- Linux 4.18.0/5.15.0
- anaconda 4.12.0

Installation

system dependencies

macOS ((install time: 15 minutes)):

  • suite-sparse (>=7.8.2)
  • graphviz (>=12.1.2)
  1. brew install suite-sparse
  2. brew install graphviz

  Manually install pygraphviz:
    export PATH=$(brew --prefix graphviz):$PATH
    export CFLAGS="-I $(brew --prefix graphviz)/include"
    export LDFLAGS="-L $(brew --prefix graphviz)/lib"
    pip install pygraphviz

debian (install time: 15 seconds):

  • libsuitesparse-dev (>=1:5.10)
  • graphviz (>=2.42.2)
  • libgraphviz-dev (>=2.42.2)
  1. apt install libsuitesparse-dev
  2. apt install graphviz libgraphviz-dev

conda (install time: 1.5 minutes):

  • conda-forge::suitesparse(>=5.10.1)
  • graphviz (>=7.1.0)
  • pygraphviz (>=1.11)
  1. conda install conda-forge::suitesparse
  2. conda install conda-forge::python-graphviz

Python dependencies

We have tested python version 3.9.0, 3.10.8, 3.10.14, 3.11.0, 3.11.5, 3.12.0.

  • python (>=3.9.0)
  • numpy (>=1.23.5)
  • matplotlib (>=3.9.1)
  • seaborn (>=0.13.2)
  • networkx (>=2.8.8)
  • pydot (>=1.4.2)
  • igraph (>=0.10.5)
  • scikit-learn (>=1.5.1)
  • scipy (>=1.14.0)
  • pandas (>=2.2.3)
  • plotly (>=5.23.0)
  • tqdm (>=4.65.0)
  • leidenalg (>= 0.9.1)
  • python-louvain (>=0.16)
  • colorcet(>=3.0.1)
  • umap-learn (>=0.5.5)
  • scikit-sparse (>=0.4.8)
  • scanpy (>=1.9.3)
  • adjustText (>=0.8)
  • pygraphviz (>=1.11)
  • gudhi (>=3.10.1)
  • magic-impute (>=3.0.0)
  • anndata (>=0.9.2)

install from pypi (0.1.5)

Expect install time on a normal computer(Intel i5-10400 (12) @ 4.300GHz): 2 minutes

pip install phlowerpy

install newest phlower version

git clone https://github.com/CostaLab/phlower.git
cd phlower
pip install .
import phlower

Tutorial

Demo

A small scRNA-seq data Fibroblast to Neuron in 5 minutes for a computer(Intel i5-10400 (12) @ 4.300GHz).

Kidney in the paper

10X multiome data Kidney.

Regulators detection example please check Regulators.

Data info

The processed data have been deposited at: DOI.

reproducibility

https://github.com/CostaLab/phlower-reproducibility

Reference

PHLOWER has been published in Nature Methods: https://doi.org/10.1101/2024.10.01.613179

Citation

@article{cheng2025phlower,
  title={PHLOWER leverages single-cell multimodal data to infer complex, multi-branching cell differentiation trajectories},
  author={Cheng, Mingbo and Jansen, Jitske and Reimer, Katharina C and Grande, Vincent P and Nagai, James S and Li, Zhijian and Kie{\ss}ling, Paul and Grasshoff, Martin and Kuppe, Christoph and Schaub, Michael T and others},
  journal={Nature Methods},
  pages={1--9},
  year={2025},
  publisher={Nature Publishing Group US New York}
}