Abstract
Maximum diversification of data is a central theme in building generalized and accurate machine learning (ML) models. In chemistry, ML has been used to develop models for predicting molecular properties, for example quantum mechanics (QM) calculated potential energy surfaces and atomic charge models. The ANI-1x and ANI-1ccx ML-based general-purpose potentials for organic molecules were developed through active learning; an automated data diversification process. Here, we describe the ANI-1x and ANI-1ccx data sets. To demonstrate data diversity, we visualize it with a dimensionality reduction scheme, and contrast against existing data sets. The ANI-1x data set contains multiple QM properties from 5 M density functional theory calculations, while the ANI-1ccx data set contains 500 k data points obtained with an accurate CCSD(T)/CBS extrapolation. Approximately 14 million CPU core-hours were expended to generate this data. Multiple QM calculated properties for the chemical elements C, H, N, and O are provided: energies, atomic forces, multipole moments, atomic charges, etc. We provide this data to the community to aid research and development of ML models for chemistry.
Highlights
Background & SummaryMachine learning (ML) and data driven methods have far reaching applications across much of science and engineering
Robotics has spearheaded the efforts to build such data sets through the use of active learning[2]: building data sets by asking machine learning (ML) models to choose what data needs to be added to a training set to perform better time
The data for training such highly flexible ML models must contain the necessary information for predicting a complete potential energy surface for a class of molecules
Summary
Machine learning (ML) and data driven methods have far reaching applications across much of science and engineering. The data for training such highly flexible ML models must contain the necessary information for predicting a complete potential energy surface for a class of molecules Constructing these data sets represents a challenging case since the dimensionality of the problem is very high, and the conformational space of a molecule is not known a priori. The ANI-1 data set[43], which consists of over 20 million conformations derived from 57 thousand distinct molecular configurations containing the C, H, N, and O chemical elements, is an example of such a data set This data set was used to build the general-purpose ANI-1 potential, which was shown to accurately predict the non-equilibrium potential surface of molecules much larger than those included in the training data set. Future extensions of these data sets will add new chemical elements and more molecular diversity will be released to the community[45]
Sign-in/Register to view highlights & summary Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
