Metadata-Version: 2.4
Name: atompaint
Version: 0.1.0
Summary: Design macromolecular interactions by in-painting full-atom models.
Author-email: Kale Kundert <kale@thekunderts.net>
Requires-Python: ~=3.10
Description-Content-Type: text/markdown
Classifier: Programming Language :: Python :: 3
License-File: LICENSE.txt
Requires-Dist: blosum
Requires-Dist: einops
Requires-Dist: escnn_atompaint
Requires-Dist: fastdigest
Requires-Dist: lightning
Requires-Dist: macromol_dataframe
Requires-Dist: macromol_gym_pretrain
Requires-Dist: macromol_gym_unsupervised
Requires-Dist: macromol_voxelize
Requires-Dist: more_itertools
Requires-Dist: multipartial
Requires-Dist: numpy
Requires-Dist: pipeline_func
Requires-Dist: polars >= 1.17
Requires-Dist: scipy
Requires-Dist: torch
Requires-Dist: torchmetrics
Requires-Dist: torchyield
Requires-Dist: tqdm
Requires-Dist: tquiet
Requires-Dist: visible_residues
Requires-Dist: xxhash
Requires-Dist: sphinx ; extra == "doc"
Requires-Dist: sphinx_rtd_theme ; extra == "doc"
Requires-Dist: autoclasstoc ; extra == "doc"
Requires-Dist: pytest ; extra == "test"
Requires-Dist: pytest_tmp_files ; extra == "test"
Requires-Dist: pytest_unordered ; extra == "test"
Requires-Dist: parametrize_from_file ; extra == "test"
Requires-Dist: hypothesis ; extra == "test"
Requires-Dist: matplotlib ; extra == "test"
Requires-Dist: colorcet ; extra == "test"
Requires-Dist: torchtest ; extra == "test"
Project-URL: Bug Tracker, https://github.com/kalekundert/atompaint/issues
Project-URL: Continuous Integration, https://github.com/kalekundert/atompaint/actions
Project-URL: Documentation, https://atompaint.readthedocs.io/en/latest/
Project-URL: Test Coverage, https://coveralls.io/github/kalekundert/atompaint
Project-URL: Version Control, https://github.com/kalekundert/atompaint
Provides-Extra: doc
Provides-Extra: test

AtomPaint
=========

[![Last release](https://img.shields.io/pypi/v/atompaint.svg)](https://pypi.python.org/pypi/atompaint)
[![Python version](https://img.shields.io/pypi/pyversions/atompaint.svg)](https://pypi.python.org/pypi/atompaint)
[![Documentation](https://img.shields.io/readthedocs/atompaint.svg)](https://torch-fuel.readthedocs.io/en/latest/)
[![Test status](https://img.shields.io/github/actions/workflow/status/kalekundert/atompaint/test.yml?branch=master)](https://github.com/kalekundert/atompaint/actions)
[![Test coverage](https://img.shields.io/codecov/c/github/kalekundert/atompaint)](https://app.codecov.io/github/kalekundert/atompaint)
[![Last commit](https://img.shields.io/github/last-commit/kalekundert/atompaint?logo=github)](https://github.com/kalekundert/atompaint)

*AtomPaint* is a collection of convolutional neural networks (CNNs) meant for 
learning from 3D images of macromolecular structures.  A unique, unifying 
feature of many of these networks is that maintain rotational equivariance; 
that is, they will recognize features no matter what orientation they appear 
in.  The models are implemented using the [PyTorch](https://pytorch.org) and 
[ESCNN](https://quva-lab.github.io/escnn/) frameworks.

Below are links to some of the datasets that these models can be trained on:

- [Macromolecular Gym](https://github.com/kalekundert/macromol_gym)
- [Atom3D](https://www.atom3d.ai/)

