Welcome to the ChemML’s documentation!

ChemML is a machine learning and informatics program suite for the analysis, mining, and modeling of chemical and materials data.

• source repository on github: https://github.com/hachmannlab/chemml

Code Design:

ChemML is developed in the Python 3 programming language and makes use of a host of data analysis and ML libraries(accessible through the Anaconda distribution), as well as domain-specific libraries. The development follows a strictly modular and object-oriented design to make the overall code as flexible and versatile as possible.

The format of library is similar to the well known libraries like Scikit-learn.

Latest Version:

• to find out about the latest version and release history, click here: https://pypi.org/project/chemml/#history

Installation and Dependencies:

We strongly recommend you to install ChemML in an Anaconda environment. The instructions to create the environment, install ChemML’s dependencies, and subsequently install Chemml using the Python Package Index (PyPI) via pip are as follows:

conda create --name chemml_env python=3.8
source activate chemml_env
conda install -c conda-forge openbabel rdkit nb_conda_kernels python-graphviz
pip install chemml

Note: Here is a list of external libraries that will be installed with chemml:

• numpy

• pandas

• tensorflow

• scikit-learn

• matplotlib

• seaborn

• lxml

• graphviz

• openpyxl

• ipywidgets

Errors during installation:

After installing ChemML in an Anaconda environment, you may run into the following error with respect to the rdkit module:

ImportError: /lib64/libstdc++.so.6: version `CXXABI_1.3.9' not found

Solution: Activate the environment in which you have installed ChemML and use the following code:

export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:/path/to/chemml_env/lib

Citation:

Please cite the use of ChemML as:

Main citation:

@article{chemml2019,
author = {Haghighatlari, Mojtaba and Vishwakarma, Gaurav and Altarawy, Doaa and Subramanian, Ramachandran and Kota, Bhargava Urala and Sonpal, Aditya and Setlur, Srirangaraj and Hachmann, Johannes},
journal = {ChemRxiv},
pages = {8323271},
title = {ChemML: A Machine Learning and Informatics Program Package for the Analysis, Mining, and Modeling of Chemical and Materials Data},
doi = {10.26434/chemrxiv.8323271.v1},
year = {2019}
}


Other references:

@article{chemml_review2019,
author = {Haghighatlari, Mojtaba and Hachmann, Johannes},
doi = {https://doi.org/10.1016/j.coche.2019.02.009},
issn = {2211-3398},
journal = {Current Opinion in Chemical Engineering},
month = {jan},
pages = {51--57},
title = {Advances of machine learning in molecular modeling and simulation},
volume = {23},
year = {2019}
}

@article{Hachmann2018,
author = {Hachmann, Johannes and Afzal, Mohammad Atif Faiz and Haghighatlari, Mojtaba and Pal, Yudhajit},
doi = {10.1080/08927022.2018.1471692},
issn = {10290435},
journal = {Molecular Simulation},
number = {11},
pages = {921--929},
title = {Building and deploying a cyberinfrastructure for the data-driven design of chemical systems and the exploration of chemical space},
volume = {44},
year = {2018}
}

@article{vishwakarma2019towards,
title={Towards autonomous machine learning in chemistry via evolutionary algorithms},
author={Vishwakarma, Gaurav and Haghighatlari, Mojtaba and Hachmann, Johannes},
journal={ChemRxiv preprint},
year={2019}
}

ChemML Library

• Introduction to ChemML Library
  • Citation:
• Input
  • Molecule
    • Construct a molecule using SMILES
    • Construct a molecule using 3D atomic coordinates
    • Construct a molecule using InChi
    • Visualize the molecule
    • Add or remove hydrogens
    • Convert chemical formats
    • Optimize XYZ geometry
• Represent
  • Feature Representation Methods in ChemML
    • Creating chemml.chem.Molecule object from molecule SMILES
• Prepare
  • Prepare: Handling Missing Values
    • ChemML implements 4 strategies to handle missing values and interpolate, replace or remove them.
    • Strategy 1: Ignoring Rows
    • Strategy 2: Replacing With Zeros
    • Strategy 3: Interpolate
    • Strategy 3: Ignore Columns
  • Prepare: Remove Constant Columns and Outliers
    • Removing Constant Columns
    • Removing oultiers based on mean
    • Removing outliers based on median
• Model
  • Build a Neural Network using ChemML
    • Fit the chemml.model.MLP model to the training data
    • Saving the ChemML model
    • Loading the saved ChemML model
    • Predict the densities for the test data
    • Evaluate model performance using chemml.utils.regression_metrics
    • Plot actual vs. predicted values
    • If the Underlying (TensorFlow/PyTorch) model is required …
    • Sometimes you may need the keras model without the output layer (for e.g., for transfer learning)
  • Neural Fingerprints
    • Splitting and preprocessing the data
    • Building the Neural Fingerprints
    • Building and training the neural network
• Optimize
  • Hyperparameter Optimization using chemml.optimization.GeneticAlgorithm
    • Defining hyperparameter space
    • Defining objective function
    • Optimize the model
    • Resume optimization
  • Feature Selection using chemml.optimization.GeneticAlgorithm
    • Defining hyperparameter space
    • Defining objective function
    • Optimize the feature space
  • Active Learning
    • Keras model example
    • Initialize active learning: EMC
    • Start the active search
    • Get a baseline: random sampling
    • Visualize distributions
    • Search in a loop
    • Create a video
• Visualize
  • Visualize
• AutoML
  • Load your data
  • Run autoML for a regression task
  • Save scores to csv

ChemML Wrapper

• Introduction to ChemML Wrapper
  • Step #1: prepare an input file
  • Step #2: run ChemML Wrapper
    • method one: Running in the Terminal
    • method two: Running in any Python Interpreter
  • Step #3: check the output directory
• ChemML Wrapper Tutorial
  • Input File Manual
    • Input File Overview
    • Specific Characters
    • General Rules
  • Input File GUI
  • Generate Morgan fingerprints from SMILES codes
  • Pyscript Usage
  • Build a simple ML model
  • Genetic Algorithm GUI tutorial
    • evaluate
    • error_metric
    • single_obj
    • space
    • test_hyperparameters

Published Models

• Published
  • Accelerated Discovery of High-Refractive-Index Molecules
    • Single Target DNNs
    • Physics Infused DNN
    • Transfer learning TZVP Polarizabilities
    • Citation

ChemML API

• ChemML API
  • Library API documentation
    • Chem module
    • Magpie_Python module
    • Initialization module
    • Datasets module
    • Preprocessing module
    • Models module
    • Optimization module
    • Visualization module
  • Wrapper API documentation
    • Table of Contents
    • Wrapper Reference

License:

ChemML is copyright (C) 2014-2022 Johannes Hachmann, Mojtaba Haghighatlari, Aditya Sonpal, Gaurav Vishwakarma and Aatish Pradhan all rights reserved. ChemML is distributed under 3-Clause BSD License (https://opensource.org/licenses/BSD-3-Clause).

About us:

Maintainers

• Johannes Hachmann, hachmann@buffalo.edu

• Aatish Pradhan, aatishpr@buffalo.edu

• Aditya Sonpal, adityaso@buffalo.edu

• Gaurav Vishwakarma

• Mojtaba Haghighatlari

University at Buffalo - The State University of New York (UB)

Contributors

• Doaa Altarawy (MolSSI): scientific advice and software mentor

• Gaurav Vishwakarma (UB): automated model optimization

• Ramachandran Subramanian (UB): Magpie descriptor library port

• Bhargava Urala Kota (UB): library database

• Aditya Sonpal (UB): graph convolution NNs

• Srirangaraj Setlur (UB): scientific advice

• Venugopal Govindaraju (UB): scientific advice

• Krishna Rajan (UB): scientific advice

• Aatish Pradhan (UB): AutoML and Jupyter GUI developer

• We encourage any contributions and feedback. Feel free to fork and make pull-request to the “development” branch.

Acknowledgements

• ChemML is based upon work supported by the U.S. National Science Foundation under grant #OAC-1751161 and in part by #OAC-1640867.

• ChemML was also supported by start-up funds provided by UB’s School of Engineering and Applied Science and UB’s Department of Chemical and Biological Engineering, the New York State Center of Excellence in Materials Informatics through seed grant #1140384-8-75163, and the U.S. Department of Energy under grant #DE-SC0017193.

• Mojtaba Haghighatlari received 2018 Phase-I and 2019 Phase-II Software Fellowships by the Molecular Sciences Software Institute (MolSSI) for his work on ChemML.