Dev/feat/load data from archive

.codecov.yml

@@ -1,4 +1,3 @@
 ignore:
   - "serenityff/charge/data"
   - "serenityff/charge/examples"
-  - "serenityff/charge/_version.py"

.github/workflows/CI.yaml

@@ -44,15 +44,15 @@ jobs:
     - name: Install package
       shell: bash -l {0}
       run: |
-        python -m pip install . --no-deps
+        pip install . --no-deps
         conda list
 
 
     - name: Run tests
       shell: bash -l {0}
       run: |
-        python setup.py install
-        pytest -v --color=yes serenityff/charge/tests/ --cov=serenityff --cov-report=xml
+        pip install . --no-deps
+        pytest -v --color=yes --cov=serenityff --cov-report=xml --run-slow serenityff/charge/tests/ tests/
 
   analyze:
     name: Analyze

.gitignore

@@ -133,3 +133,5 @@ dmypy.json
 
 # SerenityFF charge dev files
 dev
+serenityff/charge/data/additional_data/**
+!serenityff/charge/data/additional_data/.gitkeep

.pre-commit-config.yaml

@@ -44,7 +44,6 @@ repos:
           # Ignore imports in init files
           "--per-file-ignores=
             */__init__.py:F401,setup.py:E121
-            versioneer.py:W605
             ",
           # ignore long comments (E501), as long lines are formatted by black
           # ignore Whitespace before ':' (E203)

MANIFEST.in

@@ -1,4 +1,2 @@
 include LICENSE
 include MANIFEST.in
-include versioneer.py
-include serenityff/_version.py

README.md

@@ -3,8 +3,9 @@ Welcome to DASH
 
 [//]: # (Badges)
 [![CI](https://github.com/rinikerlab/DASH-tree/actions/workflows/CI.yaml/badge.svg)](https://github.com/rinikerlab/DASH-tree/actions/workflows/CI.yaml)
-[![pre-commit](https://github.com/rinikerlab/DASH-tree/actions/workflows/pre-commit.yml/badge.svg?branch=main)](https://github.com/rinikerlab/DASH-tree/actions/workflows/pre-commit.yml)
-[![arXiv](https://img.shields.io/badge/arXiv-2305.15981-b31b1b.svg)](https://doi.org/10.48550/arXiv.2305.15981)
+[![pre-commit](https://img.shields.io/badge/pre--commit-enabled-brightgreen?logo=pre-commit)](https://github.com/pre-commit/pre-commit)
+[![arXiv DASH](https://img.shields.io/badge/arXiv:DASH-2305.15981-b31b1b.svg)](https://doi.org/10.48550/arXiv.2305.15981)
+[![arXiv Props](https://img.shields.io/badge/chemrxiv:Props-2024_0ks0p-b31b1b.svg)](https://chemrxiv.org/engage/chemrxiv/article-details/6666e12112188379d8c44aa5)
 
 
 
@@ -13,18 +14,18 @@ Description
 
    Welcome to DASH. This repository is a collection of scripts, tools and other resources for partial charges in MD simulations.
 
-   It contains tools to generate partial charges for given molecules quickly, using the DASH-charge workflow and using openff to generate all other parameters. A pre computed DASH (Dynamic Attention-based Substructure Hierarchy) tree can be used to generate charges for a given molecule.
+   It contains tools to generate partial charges and other atomic and molecular properties for given molecules quickly. By using the DASH-charge workflow and using the OpenFF plugin and Forcefield to parametrize molecules quickly with QM-like charge quality. A pre-computed DASH (Dynamic Attention-based Substructure Hierarchy) tree is included and can be used to generate charges for a given molecule. Threes with different properties are available for download from the ETHZ Research Collection.
 
-   Additionally, this repository contains all tools and functions needed to generate a new decision tree for partial charge assignmend, based on the attention data of a graph neural network, cabable of predicting the partial charges of a molecule.
+   Additionally, this repository contains all tools and functions needed to generate a new DASH tree for any property, like partial charge assignments, based on the attention data of a graph neural network, capable of predicting the partial charges of a molecule.
 
-   This repository contains code for the publication by M. Lehner et al. DOI: [arXiv:2305.15981](https://doi.org/10.48550/arXiv.2305.15981) and [10.1021/acs.jcim.3c00800](https://pubs.acs.org/doi/full/10.1021/acs.jcim.3c00800)
+   This repository contains code for the publication by M. Lehner et al. DOI: [arXiv:2305.15981](https://doi.org/10.48550/arXiv.2305.15981), [10.1021/acs.jcim.3c00800](https://pubs.acs.org/doi/full/10.1021/acs.jcim.3c00800), and [chemrxiv](https://chemrxiv.org/engage/chemrxiv/article-details/6666e12112188379d8c44aa5).
 
 
 Content
 -------------
 
 * **Data Preperation**
-    * Select data from Database (ChEMBL)
+    * Select data from the Database (ChEMBL)
     * Generate diverse data set
     * Generate feature vectors
 
@@ -35,7 +36,7 @@ Content
     * Attention Extraction (GNNExplainer)
 
 * **DASH Tree**
-    * Generate DASH tree (attention based)
+    * Generate DASH tree (attention-based)
     * Tools for DASH tree
         * file I/O
         * pruning
@@ -50,7 +51,7 @@ Content
     * Tools for validation of the charges with OpenFF-Evaluator
 
 * **Examples**
-    * Examples for all important functions and tools
+    * Examples of all important functions and tools
     * A good starting point for new users
 
 
@@ -60,9 +61,44 @@ Installation
    This repository comes with a conda environment file. To install the environment, run the following command in the root directory of this repository:
 
    ```bash
-    conda env create -f environment.yml
+    conda env create -f min_environment.yml
     conda activate dash
-    python setup.py install
+    conda develop .
    ```
 
-This will create a conda enviroment with the correct packages and install the openff plugin for partial charge assignment in openff. If you plan on only using the DASH tree and not developing new trees you can also use the file `min_environment.yml` instead, which does not contain any pytorch and pytorch geometric libraries.
+This will create a conda environment with the correct packages. The environment contains the minimal dependencies required to use the tree. If the OpenFF plugin to automatically assign charge is required, use the file  `environment.yml` instead. This will create the environment, the DASH package and install the OpenFF plugin for partial charge assignment in OpenFF. The file `environment.yml` also contains all torch dependencies, used for DASH tree development.
+
+
+Usage
+-------------
+
+A default tree for MBIS partial charges is included in the repository. To use the tree, the following code can be used:
+
+```python
+# Import the DASH tree and RDKit
+from rdkit import Chem
+from serenityff.charge.tree.dash_tree import DASHTree
+# Load the default tree
+tree = DASHTree()
+# Create a RDKit molecule
+example_mol = Chem.AddHs(Chem.MolFromSmiles('CCO'))
+# Assign charges
+charges = tree.get_molecules_partial_charges(example_mol)["charges"]
+```
+
+More atomic and molecular properties can be calculated from a DASH tree populated with these properties.
+
+```python
+# Import the DASH tree and RDKit
+from rdkit import Chem
+from serenityff.charge.tree.dash_tree import DASHTree, TreeType
+from serenityff.charge.data import dash_props_tree_path
+# Load the property tree.
+# Note, that the files will be automatically downloaded the first time the tree is loaded from the ETHZ Research Collection.
+tree = DASHTree(tree_folder_path=dash_props_tree_path, tree_type=TreeType.FULL)
+# Create a RDKit molecule
+example_mol = Chem.AddHs(Chem.MolFromSmiles('CCO'))
+# Get a new property
+tree.get_property_noNAN(mol=example_mol, atom=0, property_name="DFTD4:C6")
+# Or get partial charges with a different model
+charges = tree.get_molecules_partial_charges(example_mol, chg_key="AM1BCC", chg_std_key="AM1BCC_std")["charges"]

serenityff/charge/__init__.py

@@ -1,8 +1,8 @@
 # from serenityff.charge.gnn import ChargeCorrectedNodeWiseAttentiveFP, NodeWiseAttentiveFP, Extractor, Trainer
 
-from . import _version
+# from . import _version
 
-__version__ = _version.get_versions()["version"]
+__version__ = "2.0.1"
 
 # __all__ = [
 #    Trainer,