qchem

A bleeding edge framework for molecular modeling. This framework combines the most powerful open-source libaries for machine learning, geometric deep learning, and chemoinformatics with full featured molecular datasets while remaining light weight, modular and extendable.

From Machine Learning for Molecular Simulation

In 1929 Paul Dirac stated that: “The underlying physical laws necessary for the mathematical theory of a large part of physics and the whole of chemistry are thus completely known, and the difficulty is only that the exact application of these laws leads to equations much too complicated to be soluble. It therefore becomes desirable that approximate practical methods of applying quantum mechanics should be developed, which can lead to an explanation of the main features of complex atomic systems without too much computation.” Ninety years later, this quote is still state of the art. However, in the last decade, new tools from the rapidly developing field of machine learning (ML) have started to make significant impact on the development of approximate methods for complex atomic systems, bypassing the direct solution of “equations much too complicated to be soluble”.

Datasets, models and learners implemented in pytorch

• light weight, modular, extendable
• pytorch=2.2.2, pyg=2.5.2, cuda=12.1, python=3.12.2, rdkit=2024.03.1, scikit-learn=1.4
• uses the icanswim/cosmosis datascience repo for rapid prototyping
• see setup_notes.txt for implementation details
• see experiment.ipynb for examples

References

ICLR 2021 Keynote: Geometric Deep Learning
https://youtu.be/w6Pw4MOzMuo

Geometric Deep Learning: Grids, Groups, Graphs, Geodesics, and Gauges
https://arxiv.org/abs/2104.13478

GeoMol: Torsional Geometric Generation of Molecular 3D Conformer Ensembles
https://arxiv.org/abs/2106.07802
https://github.com/PattanaikL/GeoMol

Learning Neural Generative Dynamics for Molecular Conformation Generation
https://arxiv.org/abs/2102.10240

Equivariant message passing for the prediction of tensorial properties and molecular spectra
https://arxiv.org/abs/2102.03150

Graph Neural Networks with Learnable Structural and Positional Representations
https://arxiv.org/abs/2110.07875
https://github.com/vijaydwivedi75/gnn-lspe

Physics-based Deep Learning
https://arxiv.org/abs/2109.05237

Everything is Connected: Graph Neural Networks
https://arxiv.org/abs/2301.08210

AlphaFold
https://www.nature.com/articles/s41586-021-03819-2
https://github.com/deepmind/alphafold

Deep learning for molecular design - a review of the state of the art
https://arxiv.org/abs/1903.04388

Machine learning for molecular simulation
https://arxiv.org/abs/1911.02792

Machine learning for molecular and materials science
https://www.researchgate.net/publication/326608140_Machine_learning_for_molecular_and_materials_science

Deep learning methods in protein structure prediction
https://www.sciencedirect.com/science/article/pii/S2001037019304441

Machine learning for protein folding and dynamics
https://arxiv.org/abs/1911.09811

Graph Attention Networks
https://arxiv.org/abs/1710.10903
https://youtu.be/8owQBFAHw7E

Learning to Simulate Complex Physics with Graph Networks
https://arxiv.org/abs/2002.09405
https://github.com/deepmind/deepmind-research/tree/master/learning_to_simulate

Set Transformer: A Framework for Attention-based Permutation-Invariant Neural Networks
https://arxiv.org/abs/1810.00825
https://github.com/juho-lee/set_transformer

Learning continuous and data-driven molecular descriptors by translating equivalent chemical representations
https://pubs.rsc.org/en/content/articlehtml/2019/sc/c8sc04175j

Machine Learning Force Fields
https://arxiv.org/abs/2010.07067

Unifying machine learning and quantum chemistry -- a deep neural network for molecular wavefunctions
https://arxiv.org/abs/1906.10033

Integrating Machine Learning with Physics-Based Modeling
https://arxiv.org/abs/2006.02619

Bypassing the Kohn-Sham equations with machine learning
https://www.nature.com/articles/s41467-017-00839-3

Chemistry Machine Learning Libraries

RDKit
https://www.rdkit.org/

PyG
https://pytorch-geometric.readthedocs.io/en/latest/

SchNetPack: A Deep Learning Toolbox For Atomistic Systems
https://arxiv.org/abs/1809.01072 https://github.com/atomistic-machine-learning/schnetpack

Graph Networks as a Universal Machine Learning Framework for Molecules and Crystals
https://arxiv.org/abs/1812.05055
https://github.com/materialsvirtuallab/megnet

MoleculeNet: A Benchmark for Molecular Machine Learning
https://arxiv.org/abs/1703.00564
https://github.com/deepchem/deepchem

Datasets

Quantum-Machine.org
http://quantum-machine.org/datasets/

GDB Databases
https://gdb.unibe.ch/downloads/

Alchemy: A Quantum Chemistry Dataset for Benchmarking AI Models
https://arxiv.org/abs/1906.09427
https://github.com/tencent-alchemy/Alchemy

QM7-X: A comprehensive dataset of quantum-mechanical properties spanning the chemical space of small organic molecules
https://arxiv.org/abs/2006.15139
https://zenodo.org/record/3905361

The ANI-1ccx and ANI-1x data sets, coupled-cluster and density functional theory properties for molecules
https://www.nature.com/articles/s41597-020-0473-z#Sec11

PyG Datasets
https://pytorch-geometric.readthedocs.io/en/latest/modules/datasets.html

Kaggle
https://www.kaggle.com/c/champs-scalar-coupling

QMugs: Quantum Mechanical Properties of Drug-like Molecules
https://arxiv.org/abs/2107.00367
https://doi.org/10.3929/ethz-b-000482129