Dihedral Parametrization in the Cloud with TorchANI
This is a repository where you can find a Jupyter notebook scripts to set up a protocol for parametrization of small molecules dihedrals for GAFF and OpenFF force fields using TorchANI as a reference, a PyTorch-based program for training/inference of ANI (ANAKIN-ME) deep learning models to obtain potential energy surfaces and other physical properties of molecular systems. TorchANI is open-source and freely available on GitHub: https://github.com/aiqm/torchani.
The main goal of this work is to demonstrate how to harness the power of cloud-computing to parametrize compounds in a cheap and yet feasible fashion.
ParametrizANI 
Dihedral Parametrization with TorchANI as a reference and download the topology in AMBER, GROMACS and OpenMM format.
TorchANI_2D Two Dihedral scan with TorchANI and 3D plot of the map.
Psi4+TorchANI Dihedral scan with Psi4 and structural optimization of each conformer with TorchANI.
- If you encounter any bugs, please report the issue to https://github.com/pablo-arantes/ParametrizANI/issues
- ParametrizANI by Pablo R. Arantes (@pablitoarantes)
- We would like to thank the OpenMM team for developing an excellent and open source engine.
- We would like to thank the Psi4 team for developing an excellent and open source suite of ab initio quantum chemistry.
- We would like to thank the Roitberg team for developing the fantastic TorchANI.
- We would like to thank the Xavier Barril team for their protocol on dihedrals parametrization and for the genetic algorithm script.
- We would like to thank iwatobipen for his fantastic blog on chemoinformatics.
- Also, credit to David Koes for his awesome py3Dmol plugin.
- Finally, we would like to thank Making it rain team, Pablo R. Arantes (@pablitoarantes), Marcelo D. Polêto (@mdpoleto), Conrado Pedebos (@ConradoPedebos) and Rodrigo Ligabue-Braun (@ligabue_braun), for their amazing work.
- For TorchANI, please cite:
Gao et al. "TorchANI: A Free and Open Source PyTorch-Based Deep Learning Implementation of the ANI Neural Network Potentials."
Journal of Chemical Information and Modeling (2020) doi: 10.1021/acs.jcim.0c00451 - For OpenMM, please also cite:
Eastman et al. "OpenMM 7: Rapid development of high performance algorithms for molecular dynamics."
PLOS Computational Biology (2017) doi: 10.1371/journal.pcbi.1005659 - For Molecular Dynamics Notebook, please also cite:
Arantes et al. "Making it rain: cloud-based molecular simulations for everyone."
Journal of Chemical Information and Modeling (2021) doi: 10.1021/acs.jcim.1c00998