Data

Tools

Indexes

Applications

Experiments

Open Source Science

https://github.com/adicksonlab/classicalgsg

A new logP predictor method using GSG and NNs

https://github.com/adicksonlab/classicalgsg

Science Score: 23.0%

This score indicates how likely this project is to be science-related based on various indicators:

  • ○
    CITATION.cff file
  • ○
    codemeta.json file
  • ○
    .zenodo.json file
  • ✓
    DOI references
    Found 4 DOI reference(s) in README
  • ✓
    Academic publication links
    Links to: arxiv.org, wiley.com
  • ○
    Academic email domains
  • ○
    Institutional organization owner
  • ○
    JOSS paper metadata
  • ○
    Scientific vocabulary similarity
    Low similarity (10.6%) to scientific vocabulary

Keywords

geommetric-scattering-for-graphs logp neural-networks
Last synced: 10 months ago · JSON representation

Repository

A new logP predictor method using GSG and NNs

Basic Info
  • Host: GitHub
  • Owner: ADicksonLab
  • License: mit
  • Language: Python
  • Default Branch: master
  • Homepage:
  • Size: 79.5 MB
Statistics
  • Stars: 8
  • Watchers: 4
  • Forks: 1
  • Open Issues: 1
  • Releases: 0
Topics
geommetric-scattering-for-graphs logp neural-networks
Created over 5 years ago · Last pushed over 2 years ago
Metadata Files
Readme License

README.org 

* ClassicalGSG: Prediction of logP Using Classical Molecular Dynamics Atomic Attributes and  Geometric Scattering Graphs

This project is the implementation of a method called ~ClassicalGSG~
and has introduced in [[https://onlinelibrary.wiley.com/doi/abs/10.1002/jcc.26519][ClassicalGSG: Prediction of logP using classical
molecular force fields and geometric scattering for graphs]].  In this
project, we aim to predict the partition coefficient value for the
small molecules.

Here, we use molecular features generated using a recently developed
method called [[https://arxiv.org/abs/1810.03068][Geometric Scattering for Graphs]] (GSG).  The GSG method
uses the graph structure of molecules to transform atomic attributes
into index-variant molecular features.

The atomic attributes are generated using two classical molecular
dynamics force fields generator tool [[https://cgenff.umaryland.edu][CGenFF]] and [[https://ambermd.org/AmberTools.php][Ambertools]].  We train
ClassicalGSG logP predictor models with neural networks (NNs) which,
are implemented using [[https://pytorch.org][PyTorch]].


* Installation

You should use conda to make a new virtual environment:

#+begin_src sh
  conda create -n myenv python=3.7
  conda activate myenv
#+end_src

Currently you must manually install some of the dependencies using
conda. Do this first:

#+begin_src sh
  conda install -c pytorch pytorch
  conda install -c conda-forge openbabel rdkit
#+end_src


To install from pip:

#+BEGIN_SRC bash
  pip install classicalgsg
#+END_SRC


You can install from the git repo as well:

#+begin_src sh
pip install git+https://github.com/ADicksonLab/ClassicalGSG.git
#+end_src

# TODO: inform about customizing pytorch installation

* Usage

To use our CGenFF ClassicalGSG logP predictor model run the following command:

#+BEGIN_SRC bash
 python -m LogpPredictor_CGenFF [molecule.mol2] [molecule.str]
#+END_SRC

To use our MMFF94 ClassicalGSG logP predictor model run the following command:

#+BEGIN_SRC bash
 python -m LogpPredictor_MMFF94  ['smiles']
#+END_SRC

You can generate CGenFF parameter files for your molecule using [[https://cgenff.umaryland.edu][CGenFF]]
online server.

* Dataset
The logP dataset can be downloaded from [[https://doi.org/10.5281/zenodo.4531015][Zenodo]].
Zenodo DOI: 10.5281/zenodo.4531015

Owner

  • Name: ADicksonLab
  • Login: ADicksonLab
  • Kind: organization

GitHub Events

Total
Last Year

Issues and Pull Requests

Last synced: almost 2 years ago

All Time
  • Total issues: 1
  • Total pull requests: 2
  • Average time to close issues: N/A
  • Average time to close pull requests: 2 days
  • Total issue authors: 1
  • Total pull request authors: 1
  • Average comments per issue: 8.0
  • Average comments per pull request: 0.5
  • Merged pull requests: 2
  • Bot issues: 0
  • Bot pull requests: 0
Past Year
  • Issues: 1
  • Pull requests: 2
  • Average time to close issues: N/A
  • Average time to close pull requests: 2 days
  • Issue authors: 1
  • Pull request authors: 1
  • Average comments per issue: 8.0
  • Average comments per pull request: 0.5
  • Merged pull requests: 2
  • Bot issues: 0
  • Bot pull requests: 0
View more stats
Top Authors
Issue Authors
  • fwaibl (1)
Pull Request Authors
  • ndonyapour (2)
Top Labels
Issue Labels
Pull Request Labels

Packages

  • Total packages: 1
  • Total downloads:
    • pypi 17 last-month
  • Total dependent packages: 0
  • Total dependent repositories: 1
  • Total versions: 4
  • Total maintainers: 1
pypi.org: classicalgsg

ClassicalGSG

  • Versions: 4
  • Dependent Packages: 0
  • Dependent Repositories: 1
  • Downloads: 17 Last month
Rankings
Dependent packages count: 10.0%
Stargazers count: 19.3%
Dependent repos count: 21.8%
Average: 21.9%
Forks count: 22.7%
Downloads: 35.8%
Maintainers (1)
ndonyapour
Last synced: 10 months ago

Dependencies

requirements.in pypi
  • numpy *
  • pandas *
  • parmed *
  • scikit-learn *
  • skorch *
  • tabulate *
setup.py pypi
  • ParmEd *
  • numpy *
  • pandas *
  • scikit-learn *
  • skorch *
  • tabulate *