paper.bib

@article{SOAP,
    issn = {2375-2548},
    abstract = {Determining the stability of molecules and condensed phases is the cornerstone of atomistic modeling, underpinning our understanding of chemical and materials properties and transformations. We show that a machine-learning model, based on a local description of chemical environments and Bayesian statistical learning, provides a unified framework to predict atomic-scale properties. It captures the quantum mechanical effects governing the complex surface reconstructions of silicon, predicts the stability of different classes of molecules with chemical accuracy, and distinguishes active and inactive protein ligands with more than 99\% reliability. The universality and the systematic nature of our framework provide new insight into the potential energy surface of materials and molecules.},
    journal = {Science advances},
    pages = {e1701816--e1701816},
    volume = {3},
    publisher = {American Association for the Advancement of Science},
    number = {12},
    year = {2017},
    title = {Machine learning unifies the modeling of materials and molecules},
    doi = {10.1126/sciadv.1701816},
    copyright = {http://arxiv.org/licenses/nonexclusive-distrib/1.0},
    language = {eng},
    address = {United States},
    author = {Bartók, Albert P and De, Sandip and Poelking, Carl and Bernstein, Noam and Kermode, James R and Csányi, Gábor and Ceriotti, Michele},
    keywords = {SciAdv r-articles ; Physical Sciences ; Physics},
}

@article{GaussianMomentum,
    author = {Zaverkin, V. and Kästner, J.},
    title = {Gaussian Moments as Physically Inspired Molecular Descriptors for Accurate and Scalable Machine Learning Potentials},
    journal = {Journal of Chemical Theory and Computation},
    volume = {16},
    number = {8},
    pages = {5410-5421},
    year = {2020},
    doi = {10.1021/acs.jctc.0c00347},
    note = {PMID: 32672968},
    URL = {https://doi.org/10.1021/acs.jctc.0c00347},
    eprint = {https://doi.org/10.1021/acs.jctc.0c00347},
}

@article{GMP,
    author = {Lei, Xiangyun and Medford, Andrew J.},
    title = {A Universal Framework for Featurization of Atomistic Systems},
    journal = {The Journal of Physical Chemistry Letters},
    volume = {13},
    number = {34},
    pages = {7911-7919},
    year = {2022},
    doi = {10.1021/acs.jpclett.2c02100},
    note ={PMID: 35980312},
    URL = {https://doi.org/10.1021/acs.jpclett.2c02100},
    eprint = {https://doi.org/10.1021/acs.jpclett.2c02100},
}

@article{MEGNet,
    author = {Chen, Chi and Ye, Weike and Zuo, Yunxing and Zheng, Chen and Ong, Shyue Ping},
    title = {Graph Networks as a Universal Machine Learning Framework for Molecules and Crystals},
    journal = {Chemistry of Materials},
    volume = {31},
    number = {9},
    pages = {3564-3572},
    year = {2019},
    doi = {10.1021/acs.chemmater.9b01294},
    URL = {https://doi.org/10.1021/acs.chemmater.9b01294},
    eprint = {https://doi.org/10.1021/acs.chemmater.9b01294},
}

@article{CGCNN,
    issn = {0031-9007},
    abstract = {The use of machine learning methods for accelerating the design of crystalline materials usually requires manually constructed feature vectors or complex transformation of atom coordinates to input the crystal structure, which either constrains the model to certain crystal types or makes it difficult to provide chemical insights. Here, we develop a crystal graph convolutional neural networks framework to directly learn material properties from the connection of atoms in the crystal, providing a universal and interpretable representation of crystalline materials. Our method provides a highly accurate prediction of density functional theory calculated properties for eight different properties of crystals with various structure types and compositions after being trained with 10^{4} data points. Further, our framework is interpretable because one can extract the contributions from local chemical environments to global properties. Using an example of perovskites, we show how this information can be utilized to discover empirical rules for materials design.},
    journal = {Physical review letters},
    pages = {145301--145301},
    volume = {120},
    publisher = {American Physical Society (APS)},
    number = {14},
    year = {2018},
    title = {Crystal Graph Convolutional Neural Networks for an Accurate and Interpretable Prediction of Material Properties},
    doi = {10.1103/PhysRevLett.120.145301},
    copyright = {http://arxiv.org/licenses/nonexclusive-distrib/1.0},
    language = {eng},
    address = {United States},
    author = {Xie, Tian and Grossman, Jeffrey C},
    keywords = {Physics - Materials Science ; MATERIALS SCIENCE},
    organization = {Lawrence Berkeley National Lab. (LBNL), Berkeley, CA (United States). National Energy Research Scientific Computing Center (NERSC)},
}

@article{pymatgen,
    title = {Python Materials Genomics (pymatgen): A robust, open-source python library for materials analysis},
    journal = {Computational Materials Science},
    volume = {68},
    pages = {314-319},
    year = {2013},
    issn = {0927-0256},
    doi = {10.1016/j.commatsci.2012.10.028},
    url = {https://www.sciencedirect.com/science/article/pii/S0927025612006295},
    author = {Shyue Ping Ong and William Davidson Richards and Anubhav Jain and Geoffroy Hautier and Michael Kocher and Shreyas Cholia and Dan Gunter and Vincent L. Chevrier and Kristin A. Persson and Gerbrand Ceder},
    keywords = {Materials, Project, Design, Thermodynamics, High-throughput},
    abstract = {We present the Python Materials Genomics (pymatgen) library, a robust, open-source Python library for materials analysis. A key enabler in high-throughput computational materials science efforts is a robust set of software tools to perform initial setup for the calculations (e.g., generation of structures and necessary input files) and post-calculation analysis to derive useful material properties from raw calculated data. The pymatgen library aims to meet these needs by (1) defining core Python objects for materials data representation, (2) providing a well-tested set of structure and thermodynamic analyses relevant to many applications, and (3) establishing an open platform for researchers to collaboratively develop sophisticated analyses of materials data obtained both from first principles calculations and experiments. The pymatgen library also provides convenient tools to obtain useful materials data via the Materials Project’s REpresentational State Transfer (REST) Application Programming Interface (API). As an example, using pymatgen’s interface to the Materials Project’s RESTful API and phasediagram package, we demonstrate how the phase and electrochemical stability of a recently synthesized material, Li4SnS4, can be analyzed using a minimum of computing resources. We find that Li4SnS4 is a stable phase in the Li–Sn–S phase diagram (consistent with the fact that it can be synthesized), but the narrow range of lithium chemical potentials for which it is predicted to be stable would suggest that it is not intrinsically stable against typical electrodes used in lithium-ion batteries.}
}

@article{ase-paper,
    author={Ask Hjorth Larsen and Jens Jørgen Mortensen and Jakob Blomqvist and Ivano E Castelli and Rune Christensen and Marcin
    Dułak and Jesper Friis and Michael N Groves and Bjørk Hammer and Cory Hargus and Eric D Hermes and Paul C Jennings and Peter
    Bjerre Jensen and James Kermode and John R Kitchin and Esben Leonhard Kolsbjerg and Joseph Kubal and Kristen
    Kaasbjerg and Steen Lysgaard and Jón Bergmann Maronsson and Tristan Maxson and Thomas Olsen and Lars Pastewka and Andrew
    Peterson and Carsten Rostgaard and Jakob Schiøtz and Ole Schütt and Mikkel Strange and Kristian S Thygesen and Tejs
    Vegge and Lasse Vilhelmsen and Michael Walter and Zhenhua Zeng and Karsten W Jacobsen},
    title={The atomic simulation environment—a Python library for working with atoms},
    doi = {10.1088/1361-648X/aa680e},
    journal={Journal of Physics: Condensed Matter},
    volume={29},
    number={27},
    pages={273002},
    url={http://stacks.iop.org/0953-8984/29/i=27/a=273002},
    year={2017},
    abstract={The atomic simulation environment (ASE) is a software package written in the Python programming language with the aim of setting up, steering, and analyzing atomistic simulations. In ASE, tasks are fully scripted in Python. The powerful syntax of Python combined with the NumPy array library make it possible to perform very complex simulation tasks. For example, a sequence of calculations may be performed with the use of a simple ‘for-loop’ construction. Calculations of energy, forces, stresses and other quantities are performed through interfaces to many external electronic structure codes or force fields using a uniform interface. On top of this calculator interface, ASE provides modules for performing many standard simulation tasks such as structure optimization, molecular dynamics, handling of constraints and performing nudged elastic band calculations.}
}

@article{ZuoAccelerated2021,
	title = {Accelerating materials discovery with {Bayesian} optimization and graph deep learning},
	volume = {51},
	issn = {1369-7021},
	url = {https://www.sciencedirect.com/science/article/pii/S1369702121002984},
	doi = {10.1016/j.mattod.2021.08.012},
	abstract = {Machine learning (ML) models utilizing structure-based features provide an efficient means for accurate property predictions across diverse chemical spaces. However, obtaining equilibrium crystal structures typically requires expensive density functional theory (DFT) calculations, which limits ML-based exploration to either known crystals or a small number of hypothetical crystals. Here, we demonstrate that the application of Bayesian optimization with symmetry constraints using a graph deep learning energy model can be used to perform “DFT-free” relaxations of crystal structures. Using this approach to significantly improve the accuracy of ML-predicted formation energies and elastic moduli of hypothetical crystals, two novel ultra-incompressible hard MoWC2 (P63/mmc) and ReWB (Pca21) were identified and successfully synthesized via in situ reactive spark plasma sintering from screening 399,960 transition metal borides and carbides. This work addresses a critical bottleneck to accurate property predictions for hypothetical materials, paving the way to ML-accelerated discovery of new materials with exceptional properties.},
	language = {en},
	urldate = {2023-01-26},
	journal = {Materials Today},
	author = {Zuo, Yunxing and Qin, Mingde and Chen, Chi and Ye, Weike and Li, Xiangguo and Luo, Jian and Ong, Shyue Ping},
	month = dec,
	year = {2021},
	keywords = {Bayesian optimization, Deep learning, Graph neural network, Materials discovery},
	pages = {126--135},
	file = {ScienceDirect Full Text PDF:/Users/josephmontoya/Zotero/storage/RD2HZ9WN/Zuo et al. - 2021 - Accelerating materials discovery with Bayesian opt.pdf:application/pdf;ScienceDirect Snapshot:/Users/josephmontoya/Zotero/storage/FDUV895X/S1369702121002984.html:text/html},
}

@article{CollinsAccelerated2017,
	title = {Accelerated discovery of two crystal structure types in a complex inorganic phase field},
	volume = {546},
	copyright = {2017 Macmillan Publishers Limited, part of Springer Nature. All rights reserved.},
	issn = {1476-4687},
	url = {https://www.nature.com/articles/nature22374},
	doi = {10.1038/nature22374},
	abstract = {A computational tool that combines human-like chemical understanding with ab initio methods guides the compositional choice of complex five-component metallic oxides, yielding two new complex crystal structures.},
	language = {en},
	number = {7657},
	urldate = {2023-01-26},
	journal = {Nature},
	author = {Collins, C. and Dyer, M. S. and Pitcher, M. J. and Whitehead, G. F. S. and Zanella, M. and Mandal, P. and Claridge, J. B. and Darling, G. R. and Rosseinsky, M. J.},
	month = jun,
	year = {2017},
	note = {Number: 7657, Publisher: Nature Publishing Group},
	keywords = {Materials chemistry, Theoretical chemistry, Theory and computation},
	pages = {280--284},
}

@misc{amptorch,
    title={AMPTorch},
    year={2020},
    howpublished={\url{https://github.com/ulissigroup/amptorch}},
}

@article{BehlerParrinello,
    title = {Generalized Neural-Network Representation of High-Dimensional Potential-Energy Surfaces},
    author = {Behler, J\"org and Parrinello, Michele},
    journal = {Phys. Rev. Lett.},
    volume = {98},
    issue = {14},
    pages = {146401},
    numpages = {4},
    year = {2007},
    month = {4},
    publisher = {American Physical Society},
    doi = {10.1103/PhysRevLett.98.146401},
    url = {https://link.aps.org/doi/10.1103/PhysRevLett.98.146401}
}

@inproceedings{Ray,
    author = {Moritz, Philipp and Nishihara, Robert and Wang, Stephanie and Tumanov, Alexey and Liaw, Richard and Liang, Eric and Elibol, Melih and Yang, Zongheng and Paul, William and Jordan, Michael I. and Stoica, Ion},
    title = {Ray: A Distributed Framework for Emerging AI Applications},
    year = {2018},
    isbn = {9781931971478},
    publisher = {USENIX Association},
    address = {USA},
    abstract = {The next generation of AI applications will continuously interact with the environment and learn from these interactions. These applications impose new and demanding systems requirements, both in terms of performance and flexibility. In this paper, we consider these requirements and present Ray--a distributed system to address them. Ray implements a unified interface that can express both task-parallel and actor-based computations, supported by a single dynamic execution engine. To meet the performance requirements, Ray employs a distributed scheduler and a distributed and fault-tolerant store to manage the system's control state. In our experiments, we demonstrate scaling beyond 1.8 million tasks per second and better performance than existing specialized systems for several challenging reinforcement learning applications.},
    booktitle = {Proceedings of the 13th USENIX Conference on Operating Systems Design and Implementation},
    pages = {561–577},
    numpages = {17},
    location = {Carlsbad, CA, USA},
    series = {OSDI'18}
}