Metadata-Version: 2.1
Name: prescient
Version: 0.0.2
Summary: Method for simulating single cells using longitudinal scRNA-seq.
Home-page: https://github.com/gifford-lab/prescient
Author: Sachit Saksena; Grace Hui-Ting Yeo
Author-email: sachit@mit.edu
License: MIT License
Platform: UNKNOWN
Classifier: Intended Audience :: Science/Research
Classifier: License :: OSI Approved :: MIT License
Classifier: Operating System :: OS Independent
Classifier: Programming Language :: Python :: 3
Requires-Python: >=3.4
Description-Content-Type: text/markdown
Requires-Dist: scanpy (>=1.7)
Requires-Dist: pyreadr (>=0.0)
Requires-Dist: matplotlib (>=3.3)
Requires-Dist: annoy (>=1.17.0)
Requires-Dist: numpy (>=1.14)
Requires-Dist: pandas (>=0.25)
Requires-Dist: scikit-learn (>=0.21)
Requires-Dist: scipy (>=1.3)
Requires-Dist: setuptools (>=41.6)
Requires-Dist: torch (>=1.5)
Requires-Dist: torchvision (>=0.7)
Requires-Dist: geomloss (==0.2.3)
Requires-Dist: pykeops (>=1.3)

# prescient
Software for PRESCIENT (Potential eneRgy undErlying Single Cell gradIENTs), a generative model for modeling single-cell time-series.
+ Documentation available at prescient.github.io.
+ Current paper version: https://www.biorxiv.org/content/10.1101/2020.08.26.269332v1
+ For paper pre-processing scripts, training bash scripts, pre-trained models, and visualization notebooks please visit https://github.com/gifford-lab/prescient-analysis.

<!-- ![trajectories_gif](docs/assets/gifs/trajectories.gif) -->

## Requirements

+ pytorch 1.4.0
+ geomloss 0.2.3, pykeops 1.3
+ numpy, scipy, pandas, sklearn, tqdm, annoy
+ scanpy, pyreadr, anndata
+ Recommended: An Nvidia GPU with CUDA support for GPU acceleration (see paper for more details on computational resources)

## Documentation
Documentation is available at https://cgs.csail.mit.edu/prescient.


## Bugs & Suggestions

Please report any bugs, problems, suggestions or requests as a [Github issue](https://github.com/gifford-lab/prescient/issues)


