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

Ramakrishnan, Raghunathan; Dral, Pavlo O.; Rupp, Matthias; Lilienfeld, O. Anatole von

doi:10.1021/acs.jctc.5b00099

Cited by 794 publications

(930 citation statements)

References 48 publications

Supporting

Mentioning

875

Contrasting

Unclassified

Order By: Relevance

“…The finding that the ML models also improve on the baseline method's outliers agrees with conclusions drawn in a previous finding where we applied the ∆-ML Ansatz to model DFT-level enthalpies of atomization for the 134k dataset, and where we found that for the most extreme outlier the baseline model's error reduced systematically with the training set size of the augmenting ML model. 23 When considering the 10 most extreme outliers of the ML models in Table I, neither order nor identity of the DFT outliers is conserved. Among the top 10 outliers of the 5k model, for example, there is even a saturated molecule from the opposite (blue) end of the error distribution in Fig.…”

Section: Dft and ML Model Outliersmentioning

confidence: 99%

“…20 By now, ML models have been shown to reach the highly coveted quantum chemical accuracy for many different ground-state molecular properties. [21][22][23] As such, also quantum mechanical expectation values can be interpolated in chemical space. 17 Improvement of molecular models of chemical properties based on molecular similarity 24,25 is also related to this approach.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Electronic spectra from TDDFT and machine learning in chemical space

Ramakrishnan

Hartmann²,

Tapavicza³

et al. 2015

The Journal of Chemical Physics

Self Cite

264

311

View full text Add to dashboard Cite

Due to its favorable computational efficiency, time-dependent (TD) density functional theory (DFT) enables the prediction of electronic spectra in a high-throughput manner across chemical space. Its predictions, however, can be quite inaccurate. We resolve this issue with machine learning models trained on deviations of reference second-order approximate coupled-cluster (CC2) singles and doubles spectra from TDDFT counterparts, or even from DFT gap. We applied this approach to low-lying singlet-singlet vertical electronic spectra of over 20 000 synthetically feasible small organic molecules with up to eight CONF atoms. The prediction errors decay monotonously as a function of training set size. For a training set of 10 000 molecules, CC2 excitation energies can be reproduced to within ±0.1 eV for the remaining molecules. Analysis of our spectral database via chromophore counting suggests that even higher accuracies can be achieved. Based on the evidence collected, we discuss open challenges associated with data-driven modeling of high-lying spectra and transition intensities. C 2015 AIP Publishing LLC. [http://dx

show abstract

Section: Dft and ML Model Outliersmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Electronic spectra from TDDFT and machine learning in chemical space

Ramakrishnan

Hartmann²,

Tapavicza³

et al. 2015

The Journal of Chemical Physics

Self Cite

264

311

View full text Add to dashboard Cite

show abstract

“…Inspired by the high-dimensional NN reported by Behler et al, 35 we developed a neural network method named QM/MM-NN for ab initio QM/MM potential energy calculations to reduce the expensive ab initio computational cost. Similar to the Δ-machine learning method, 47 the potential energy difference between two levels was chosen as the output variable of our QM/MM-NN. We aim to develop an artificial neural network such that the NN potential energy predictions closely approximate the ab initio QM/MM calculations, and the free energy changes along the reaction coordinate through QM/MM-NN reweighting achieve similar free energy profiles.…”

Section: Introductionmentioning

confidence: 99%

Multiscale Quantum Mechanics/Molecular Mechanics Simulations with Neural Networks

Shen

Yang

2016

J. Chem. Theory Comput.

123

143

View full text Add to dashboard Cite

Molecular dynamics simulation with multiscale quantum mechanics/molecular mechanics (QM/MM) methods is a very powerful tool for understanding the mechanism of chemical and biological processes in solution or enzymes. However, its computational cost can be too high for many biochemical systems because of the large number of ab initio QM calculations. Semiempirical QM/MM simulations have much higher efficiency. Its accuracy can be improved with a correction to reach the ab initio QM/MM level. The computational cost on the ab initio calculation for the correction determines the efficiency. In this paper we developed a neural network method for QM/MM calculation as an extension of the neural-network representation reported by Behler and Parrinello. With this approach, the potential energy of any configuration along the reaction path for a given QM/MM system can be predicted at the ab initio QM/MM level based on the semiempirical QM/MM simulations. We further applied this method to three reactions in water to calculate the free energy changes. The free-energy profile obtained from the semiempirical QM/MM simulation is corrected to the ab initio QM/MM level with the potential energies predicted with the constructed neural network. The results are in excellent accordance with the reference data that are obtained from the ab initio QM/MM molecular dynamics simulation or corrected with direct ab initio QM/MM potential energies. Compared with the correction using direct ab initio QM/MM potential energies, our method shows a speed-up of 1 or 2 orders of magnitude. It demonstrates that the neural network method combined with the semiempirical QM/MM calculation can be an efficient and reliable strategy for chemical reaction simulations.

show abstract

“…However, the ability to learn patterns from the data increases the flexibility of machine learning algorithms. The technique is used in many fields, including determining disease risks resulting from DNA mutations,11 engineering sciences,12 space science,13 and chemistry 14…”

Section: Introductionmentioning

confidence: 99%

A novel prediction method for lymph node involvement in endometrial cancer: machine learning

Günakan

Atan

Haberal

et al. 2018

Int J Gynecol Cancer

View full text Add to dashboard Cite

ObjectiveThe necessity of lymphadenectomy and the prediction of lymph node involvement (LNI) in endometrial cancer (EC) have been hotly-debated questions in recent years. Machine learning is a broad field that can produce results and estimations. In this study we constructed prediction models for EC patients using the Naïve Bayes machine learning algorithm for LNI prediction.MethodsThe study assessed 762 patients with EC. Algorithm models were based on the following histopathological factors: V1: final histology; V2: presence of lymphovascular space invasion (LVSI); V3: grade; V4: tumor diameter; V5: depth of myometrial invasion (MI); V6: cervical glandular stromal invasion (CGSI); V7: tubal or ovarian involvement; and V8: pelvic LNI. Logistic regression analysis was also used to evaluate the independent factors affecting LNI.ResultsThe mean age of patients was 59.1 years. LNI was detected in 102 (13.4%) patients. Para-aortic LNI (PaLNI) was detected in 54 (7.1%) patients, of which four patients had isolated PaLNI. The accuracy rate of the algorithm models was found to be between 84.2% and 88.9% and 85.0% and 97.6% for LNI and PaLNI, respectively. In multivariate analysis, the histologic type, LVSI, depth of MI, and CGSI were independently and significantly associated with LNI (p<0.001 for all).ConclusionsMachine learning may have a place in the decision tree for the management of EC. This is a preliminary report about the use of a new statistical technique. Larger studies with the addition of sentinel lymph node status, laboratory findings, or imaging results with machine learning algorithms may herald a new era in the management of EC.

show abstract

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

Cited by 794 publications

References 48 publications

Electronic spectra from TDDFT and machine learning in chemical space

Electronic spectra from TDDFT and machine learning in chemical space

Multiscale Quantum Mechanics/Molecular Mechanics Simulations with Neural Networks

A novel prediction method for lymph node involvement in endometrial cancer: machine learning

Contact Info

Product

Resources

About