MIOFlow is a Python package for modeling and analyzing single-cell RNA-seq data using optimal flows. It leverages neural ordinary differential equations (neural ODEs) and optimal transport to reconstruct cell developmental trajectories from time-series scRNA-seq data.
Features
- Trajectory inference using optimal transport and neural ODEs
- GAGA embedding — geometric autoencoder that preserves PHATE distances in latent space
- Gene-space decoding — map trajectories back to full gene expression via PCA inverse projection
- Flexible I/O for AnnData and standard scRNA-seq formats
Installation
Install from PyPI
pip install MIOFlow
Install from GitHub (Development Version)
pip install git+https://github.com/yourusername/MIOFlow.git
Usage
Basic Workflow
from MIOFlow.gaga import Autoencoder, train_gaga_two_phase, train_valid_loader_from_pc
from MIOFlow.mioflow import MIOFlow
# 1. Train a GAGA autoencoder on PCA coordinates + PHATE distances
gaga_model = Autoencoder(input_dim=50, latent_dim=10)
train_loader, val_loader = train_valid_loader_from_pc(X_pca, D_phate, batch_size=64)
train_gaga_two_phase(gaga_model, train_loader, encoder_epochs=50, decoder_epochs=50)
# 2. Fit MIOFlow
mf = MIOFlow(
adata,
gaga_model=gaga_model,
gaga_input_scaler=scaler, # StandardScaler fitted on X_pca
obs_time_key='day',
n_epochs=200,
)
mf.fit()
# 3. Inspect results
print(mf.trajectories.shape) # (n_bins, n_trajectories, latent_dim)
gene_traj = mf.decode_to_gene_space() # (n_bins, n_trajectories, n_genes)
Full worked examples are in the tutorials/ directory.
Citation
If you use MIOFlow in your research, please cite:
@misc{https://doi.org/10.48550/arxiv.2206.14928,
doi = {10.48550/ARXIV.2206.14928},
url = {https://arxiv.org/abs/2206.14928},
author = {Huguet, Guillaume and Magruder, D. S. and Tong, Alexander and Fasina, Oluwadamilola and Kuchroo, Manik and Wolf, Guy and Krishnaswamy, Smita},
keywords = {Machine Learning (cs.LG), FOS: Computer and information sciences, FOS: Computer and information sciences},
title = {Manifold Interpolating Optimal-Transport Flows for Trajectory Inference},
publisher = {arXiv},
year = {2022},
copyright = {arXiv.org perpetual, non-exclusive license}
}
License
MIOFlow is distributed under the terms of the Yale License.
Support
- 🐛 Issues: GitHub Issues
- 💬 Discussions: GitHub Discussions
Acknowledgments
- Built with PyTorch and torchdiffeq for neural ODE integration
- Integrates with the scanpy / AnnData ecosystem for single-cell analysis
- Optimal transport via the POT library
- Geometric embedding via PHATE
Metadata
Release files for mioflow 0.1.14
For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.
Source distribution (sdist)
| File | Size | Uploaded | Provenance | Actions |
|---|---|---|---|---|
| mioflow-0.1.14.tar.gz | 20.8 kB | Without provenance | Details for mioflow-0.1.14.tar.gz |
Built distribution (wheel)
| File | Interpreter | ABI | Platform | Actions Reset |
|---|---|---|---|---|
| mioflow-0.1.14-py3-none-any.whl | Python 3 | none | any | Details for mioflow-0.1.14-py3-none-any.whl |
Total release size: 43.4 kB
Release files / mioflow-0.1.14.tar.gz
| Download URL | mioflow-0.1.14.tar.gz |
|---|---|
| Size | 20.8 kB |
| Tags | Source |
|
SHA-256 checksum How to use checksums |
736c29b9175443060ba1bac3eb7f98753414c310cccdc8f3c64fa895ffc83cb0
|
|
BLAKE2b-256 checksum How to use checksums |
453c714cb85acbfc1f43e40a80b8dfdbe4340aa8be9deef3fb3b991324d5f707
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
twine/6.1.0 CPython/3.10.17
|
Release files / mioflow-0.1.14-py3-none-any.whl
| Download URL | mioflow-0.1.14-py3-none-any.whl |
|---|---|
| Size | 22.6 kB |
| Tags | Python 3 |
|
SHA-256 checksum How to use checksums |
d067e77074b05295ce4ed3857b912fc5559c4e592e82fdb465e8bbd5684ca2a7
|
|
BLAKE2b-256 checksum How to use checksums |
f185074c485aa1429421dc9b12ec7930ab6d8df34cd174dfb31deb4a426b55bc
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
twine/6.1.0 CPython/3.10.17
|
Security for mioflow 0.1.14
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
No provenance found
None of the files in this release were uploaded with provenance metadata.
A note on package security
PyPI is a repository for distributing packages, not a service for auditing them. While third-party security researchers and our staff do their best to detect and remove malicious packages, PyPI cannot guarantee the quality, security or suitability of any given package.
To learn more about protecting yourself from malicious actors, visit our package security documentation.