Big Data Meets Quantum Chemistry Approximations: The Δ-Machine Learning Approach.

By R. Ramakrishnan, Matthias Rupp , Pavlo O. Dral, Max-Planck-Institut f¨ur Kohlenforschung, Maria Ruhr, Germany O. Anatole, von Lilienfeld. 2015

794 citations recorded by the source corpus.

Chemically accurate and comprehensive studies of the virtual space of all possible molecules are severely limited by the computational cost of quantum chemistry. We introduce a composite strategy that adds machine learning corrections to computationally inexpensive approximate legacy quantum methods. After training, highly accurate predictions of enthalpies, free energies, entropies, and electron correlation energies are possible, for significantly larger molecular sets than used for training. For thermochemical properties of up to 16k isomers of C7H10O2 we present numerical evidence that chemical accuracy can be reached. We also predict electron correlation energy in post Hartree-Fock methods, at the computational cost of Hartree-Fock, and we establish a qualitative relationship between molecular entropy and electron correlation. The transferability of our approach is demonstrated, using semiempirical quantum chemistry and machine learning models trained on 1 and 10% of 134k organic molecules, to reproduce enthalpies of all remaining molecules at density functional theory level of accuracy.

Subfields: Computational Theory and Mathematics, Materials Chemistry, Molecular Biology

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computational Theory and Mathematics

1. Tommi S. Jaakkola
2. Regina Barzilay
3. Wengong Jin
4. Sam Schoenholz
5. Alexandre Tkatchenko
6. Connor W. Coley
7. Klaus‐Robert Müller 
8. Jure Leskovec

In Materials Chemistry

1. Tommi S. Jaakkola
2. Regina Barzilay
3. Wengong Jin
4. Alexandre Tkatchenko
5. Sam Schoenholz
6. Connor W. Coley
7. Jure Leskovec
8. Klaus‐Robert Müller 

In Molecular Biology

1. Regina Barzilay
2. Tommi S. Jaakkola
3. Wengong Jin
4. Klaus‐Robert Müller 
5. Steven M. Kearnes
6. Alexandre Tkatchenko
7. O. V. Lilienfeld
8. Connor W. Coley

Citations in the corpus

• Learning a Local-Variable Model of Aromatic and Conjugated Systems (2018, methodology): “27,28 Encouraged by this success, several groups aim to replace computationally expensive ab initio calculations with accurate approximations in molecular simulations.”
• Prediction Errors of Molecular Machine Learning Models Lower than Hybrid DFT Error. (2017, background): “8 ), therefore represents an appealing alternative strategy.”
• A graph-convolutional neural network model for the prediction of chemical reactivity (2018, background): “[28] Raghunathan Ramakrishnan, Pavlo O.”