Kevin A. Spiekermann
Research Subjects
Kevin A. Spiekermann has a total of 70 co-authors. Here are the top co-authors by number of shared publications:
- William H. Green
Massachusetts Institute of Technology
12 shared publications - Haoyang Wu
Massachusetts Institute of Technology
6 shared publications - Xiaorui Dong
Department of Chemical Engineering
6 shared publications Angiras Menon
Department of Chemical Engineering, Massachusetts Institute of Technology, Cambridge, Massachusetts 02139, United States
4 shared publications- Hao-Wei Pang
Massachusetts Institute of Technology
4 shared publications - Jackson W. Burns
Massachusetts Institute of Technology 77 Massachusetts Avenue
4 shared publications
Latest Publications
- Open Access
- 30 July 2026
- Published by American Chemical Society (ACS) in ChemRxiv
QuantumPioneer: Scalable generation of quantum chemical data for solution-phase hydrogen transfer reactions
High-fidelity quantum chemical (QM) datasets that jointly resolve reaction thermochemistry, kinetics, and solvation at scale remain scarce, especially for radical chemistry. We introduce QuantumPioneer, an open-access reaction-centered QM database and workflow for small organic molecules, focused on peroxyl-mediated hydrogen atom transfer (HAT) and the corresponding homolytic bond dissociation reactions. QuantumPioneer contains 348,258 species (2–21 heavy atoms), 167,237 validated HAT transition states (TS) with corresponding reaction energies and homolytic bond dissociation energies (BDEs), and over 100 million COSMO-RS solvation free energies (∆Go_solv) and enthalpies (∆Ho_solv) across 295 solvents. The workflow uses ωB97X-D/def2-SVP geometries, DLPNO-CCSD(T)-F12d/cc-pVTZ-F12 single-point energies, empirical thermochemical corrections, transition-state theory, and COSMORS BP-TZVPD-FINE solvation in a single high-throughput pipeline. Our benchmarks show reliable accuracy, with mean absolute errors compared to experimental data of 0.82 kcal/mol for gas-phase enthalpies of formation, 1.60 kcal/mol for C–H BDEs, 1.45 kcal/mol for HAT barriers, and 0.57 kcal/mol for ∆Go_solv values. We demonstrate two predictive applications. First, we show that a combined BDE and HAT-barrier model identifies experimentally observed oxidative degradation sites in drug-like molecules with a 91% top-5 hit rate and 80% site-level recall. Second, a QM-parameterized Abraham model enables rapid solvation energy estimates at near-COSMO-RS accuracy within its training domain, reproducing ∆Go_solv and ∆Ho_solv values with MAEs of 0.16 kcal/mol and 0.18 kcal/mol, respectively, though performance on experimental ΔGo_solv values for unseen solutes was worse, with MAEs of 1.32 kcal/mol, providing a scalable template for other reaction families, unifying equilibrium species, validated TS, thermochemistry, kinetics, and solvation into one workflow.
- 30 July 2026
- Published by American Chemical Society (ACS) in ChemRxiv
Open AccessQuantumPioneer: Scalable generation of quantum chemical data for solution-phase hydrogen transfer reactions
High-fidelity quantum chemical (QM) datasets that jointly resolve reaction thermochemistry, kinetics, and solvation at scale remain scarce, especially for radical chemistry. We introduce QuantumPioneer, an open-access reaction-centered QM database and workflow for small organic molecules, focused on peroxyl-mediated hydrogen atom transfer (HAT) and the corresponding homolytic bond dissociation reactions. QuantumPioneer contains 348,258 species (2–21 heavy atoms), 167,237 validated HAT transition states (TS) with corresponding reaction energies and homolytic bond dissociation energies (BDEs), and over 100 million COSMO-RS solvation free energies (∆Go_solv) and enthalpies (∆Ho_solv) across 295 solvents. The workflow uses ωB97X-D/def2-SVP geometries, DLPNO-CCSD(T)-F12d/cc-pVTZ-F12 single-point energies, empirical thermochemical corrections, transition-state theory, and COSMORS BP-TZVPD-FINE solvation in a single high-throughput pipeline. Our benchmarks show reliable accuracy, with mean absolute errors compared to experimental data of 0.82 kcal/mol for gas-phase enthalpies of formation, 1.60 kcal/mol for C–H BDEs, 1.45 kcal/mol for HAT barriers, and 0.57 kcal/mol for ∆Go_solv values. We demonstrate two predictive applications. First, we show that a combined BDE and HAT-barrier model identifies experimentally observed oxidative degradation sites in drug-like molecules with a 91% top-5 hit rate and 80% site-level recall. Second, a QM-parameterized Abraham model enables rapid solvation energy estimates at near-COSMO-RS accuracy within its training domain, reproducing ∆Go_solv and ∆Ho_solv values with MAEs of 0.16 kcal/mol and 0.18 kcal/mol, respectively, though performance on experimental ΔGo_solv values for unseen solutes was worse, with MAEs of 1.32 kcal/mol, providing a scalable template for other reaction families, unifying equilibrium species, validated TS, thermochemistry, kinetics, and solvation into one workflow.
- Open Access
- 12 July 2026
- Published by American Chemical Society (ACS) in ChemRxiv
QuantumPioneer: Scalable generation of quantum chemical data for solution-phase hydrogen transfer reactions
High-fidelity quantum chemical (QM) datasets that jointly resolve reaction thermochemistry, kinetics, and solvation at scale remain scarce, especially for radical chemistry. We introduce QuantumPioneer, an open-access reaction-centered QM database and workflow for small organic molecules, focused on peroxyl-mediated hydrogen atom transfer (HAT) and the corresponding homolytic bond dissociation reactions. QuantumPioneer contains 348,258 species (2–21 heavy atoms), 167,237 validated HAT transition states (TS) with corresponding reaction energies and homolytic bond dissociation energies (BDEs), and over 100 million COSMO-RS solvation free energies (∆Go_solv) and enthalpies (∆Ho_solv) across 295 solvents. The workflow uses ωB97X-D/def2-SVP geometries, DLPNO-CCSD(T)-F12d/def2-TZVP single-point energies, empirical thermochemical corrections, transition-state theory, and COSMORS BP-TZVPD-FINE solvation in a single high-throughput pipeline. Our benchmarks show reliable accuracy, with mean absolute errors compared to experimental data of 0.82 kcal/mol for gas-phase enthalpies of formation, 1.60 kcal/mol for C–H BDEs, 1.45 kcal/mol for HAT barriers, and 0.57 kcal/mol for ∆Go_solv values. We demonstrate two predictive applications. First, we show that a combined BDE and HAT-barrier model identifies experimentally observed oxidative degradation sites in drug-like molecules with a 91% top-5 hit rate and 80% site-level recall. Second, a QM-parameterized Abraham model enables rapid solvation energy estimates at near-COSMO-RS accuracy within its training domain, reproducing ∆Go_solv and ∆Ho_solv values with MAEs of 0.16 kcal/mol and 0.18 kcal/mol, respectively, though performance on experimental ΔGo_solv values for unseen solutes was worse, with MAEs of 1.32 kcal/mol, providing a scalable template for other reaction families, unifying equilibrium species, validated TS, thermochemistry, kinetics, and solvation into one workflow.
- 12 July 2026
- Published by American Chemical Society (ACS) in ChemRxiv
Open AccessQuantumPioneer: Scalable generation of quantum chemical data for solution-phase hydrogen transfer reactions
High-fidelity quantum chemical (QM) datasets that jointly resolve reaction thermochemistry, kinetics, and solvation at scale remain scarce, especially for radical chemistry. We introduce QuantumPioneer, an open-access reaction-centered QM database and workflow for small organic molecules, focused on peroxyl-mediated hydrogen atom transfer (HAT) and the corresponding homolytic bond dissociation reactions. QuantumPioneer contains 348,258 species (2–21 heavy atoms), 167,237 validated HAT transition states (TS) with corresponding reaction energies and homolytic bond dissociation energies (BDEs), and over 100 million COSMO-RS solvation free energies (∆Go_solv) and enthalpies (∆Ho_solv) across 295 solvents. The workflow uses ωB97X-D/def2-SVP geometries, DLPNO-CCSD(T)-F12d/def2-TZVP single-point energies, empirical thermochemical corrections, transition-state theory, and COSMORS BP-TZVPD-FINE solvation in a single high-throughput pipeline. Our benchmarks show reliable accuracy, with mean absolute errors compared to experimental data of 0.82 kcal/mol for gas-phase enthalpies of formation, 1.60 kcal/mol for C–H BDEs, 1.45 kcal/mol for HAT barriers, and 0.57 kcal/mol for ∆Go_solv values. We demonstrate two predictive applications. First, we show that a combined BDE and HAT-barrier model identifies experimentally observed oxidative degradation sites in drug-like molecules with a 91% top-5 hit rate and 80% site-level recall. Second, a QM-parameterized Abraham model enables rapid solvation energy estimates at near-COSMO-RS accuracy within its training domain, reproducing ∆Go_solv and ∆Ho_solv values with MAEs of 0.16 kcal/mol and 0.18 kcal/mol, respectively, though performance on experimental ΔGo_solv values for unseen solutes was worse, with MAEs of 1.32 kcal/mol, providing a scalable template for other reaction families, unifying equilibrium species, validated TS, thermochemistry, kinetics, and solvation into one workflow.
- Open Access
- 18 May 2026
- Published by American Chemical Society (ACS) in ChemRxiv
QuantumPioneer: Scalable generation of quantum chemical data for solution-phase hydrogen transfer reactions
High-fidelity quantum chemical (QM) datasets that jointly resolve reaction thermochemistry, kinetics, and solvation at scale remain scarce, especially for radical chemistry. We introduce QuantumPioneer, an open-access reaction-centered QM database and workflow for small organic molecules, focused on peroxyl-mediated hydrogen atom transfer (HAT) and the corresponding homolytic bond dissociation reactions. QuantumPioneer contains 348,258 species (2–21 heavy atoms), 167,237 validated HAT transition states (TS) with corresponding reaction energies and homolytic bond dissociation energies (BDEs), and over 100 million COSMO-RS solvation free energies (∆Go solv) and enthalpies (∆Ho solv) across 295 solvents. The workflow uses ωB97X-D/def2-SVP geometries, DLPNO-CCSD(T)-F12d/def2-TZVP single-point energies, empirical thermochemical corrections, transition-state theory, and COSMORS BP-TZVPD-FINE solvation in a single high-throughput pipeline. Our benchmarks show reliable accuracy, with mean absolute errors compared to experimental data of 0.82 kcal/mol for gas-phase enthalpies of formation, 1.60 kcal/mol for C–H BDEs, 1.45 kcal/mol for HAT barriers, and 0.57 kcal/mol for ∆Go solv values. We demonstrate two predictive applications. First, we show that a combined BDE and HAT-barrier model identifies experimentally observed oxidative degradation sites in drug-like molecules with a 91% top-5 hit rate and 80% site-level recall. Second, a QM-parameterized Abraham model enables rapid solvation energy estimates at near-COSMO-RS accuracy within its training domain, reproducing ∆Go solv and ∆Ho solv
- 18 May 2026
- Published by American Chemical Society (ACS) in ChemRxiv
Open AccessQuantumPioneer: Scalable generation of quantum chemical data for solution-phase hydrogen transfer reactions
High-fidelity quantum chemical (QM) datasets that jointly resolve reaction thermochemistry, kinetics, and solvation at scale remain scarce, especially for radical chemistry. We introduce QuantumPioneer, an open-access reaction-centered QM database and workflow for small organic molecules, focused on peroxyl-mediated hydrogen atom transfer (HAT) and the corresponding homolytic bond dissociation reactions. QuantumPioneer contains 348,258 species (2–21 heavy atoms), 167,237 validated HAT transition states (TS) with corresponding reaction energies and homolytic bond dissociation energies (BDEs), and over 100 million COSMO-RS solvation free energies (∆Go solv) and enthalpies (∆Ho solv) across 295 solvents. The workflow uses ωB97X-D/def2-SVP geometries, DLPNO-CCSD(T)-F12d/def2-TZVP single-point energies, empirical thermochemical corrections, transition-state theory, and COSMORS BP-TZVPD-FINE solvation in a single high-throughput pipeline. Our benchmarks show reliable accuracy, with mean absolute errors compared to experimental data of 0.82 kcal/mol for gas-phase enthalpies of formation, 1.60 kcal/mol for C–H BDEs, 1.45 kcal/mol for HAT barriers, and 0.57 kcal/mol for ∆Go solv values. We demonstrate two predictive applications. First, we show that a combined BDE and HAT-barrier model identifies experimentally observed oxidative degradation sites in drug-like molecules with a 91% top-5 hit rate and 80% site-level recall. Second, a QM-parameterized Abraham model enables rapid solvation energy estimates at near-COSMO-RS accuracy within its training domain, reproducing ∆Go solv and ∆Ho solv
- 19 September 2024
- Published by American Chemical Society (ACS) in The Journal of Physical Chemistry A
- Vol. 128 (39) , 8384-8403
Accurately Predicting Barrier Heights for Radical Reactions in Solution Using Deep Graph Networks
Quantitative estimates of reaction barriers and solvent effects are essential for developing kinetic mechanisms and predicting reaction outcomes. Here, we create a new data set of 5,600 unique elementary radical reactions calculated using the M06-2X/def2-QZVP//B3LYP-D3(BJ)/def2-TZVP level of theory. A conformer search is done for each species using TPSS/def2-TZVP. Gibbs free energies of activation and of reaction for these radical reactions in 40 common solvents are obtained using COSMO-RS for solvation effects. These balanced reactions involve the elements H, C, N, O, and S, contain up to 19 heavy atoms, and have atom-mapped SMILES. All transition states are verified by an intrinsic reaction coordinate calculation. We next train a deep graph network to directly estimate the Gibbs free energy of activation and of reaction in both gas and solution phases using only the atom-mapped SMILES of the reactant and product and the SMILES of the solvent. This simple input representation avoids computationally expensive optimizations for the reactant, transition state, and product structures during inference, making our model well-suited for high-throughput predictive chemistry and quickly providing information for (retro-)synthesis planning tools. To properly measure model performance, we report results on both interpolative and extrapolative data splits and also compare to several baseline models. During training and testing, the data set is augmented by including the reverse direction of each reaction and variants with different resonance structures. After data augmentation, we have around 2 million entries to train the model, which achieves a testing set mean absolute error of 1.16 kcal mol-1 for the Gibbs free energy of activation in solution. We anticipate this model will accelerate predictions for high-throughput screening to quickly identify relevant reactions in solution, and our data set will serve as a benchmark for future studies.
- 19 September 2024
- Published by American Chemical Society (ACS) in The Journal of Physical Chemistry A
- Vol. 128 (39) , 8384-8403
Accurately Predicting Barrier Heights for Radical Reactions in Solution Using Deep Graph Networks
Quantitative estimates of reaction barriers and solvent effects are essential for developing kinetic mechanisms and predicting reaction outcomes. Here, we create a new data set of 5,600 unique elementary radical reactions calculated using the M06-2X/def2-QZVP//B3LYP-D3(BJ)/def2-TZVP level of theory. A conformer search is done for each species using TPSS/def2-TZVP. Gibbs free energies of activation and of reaction for these radical reactions in 40 common solvents are obtained using COSMO-RS for solvation effects. These balanced reactions involve the elements H, C, N, O, and S, contain up to 19 heavy atoms, and have atom-mapped SMILES. All transition states are verified by an intrinsic reaction coordinate calculation. We next train a deep graph network to directly estimate the Gibbs free energy of activation and of reaction in both gas and solution phases using only the atom-mapped SMILES of the reactant and product and the SMILES of the solvent. This simple input representation avoids computationally expensive optimizations for the reactant, transition state, and product structures during inference, making our model well-suited for high-throughput predictive chemistry and quickly providing information for (retro-)synthesis planning tools. To properly measure model performance, we report results on both interpolative and extrapolative data splits and also compare to several baseline models. During training and testing, the data set is augmented by including the reverse direction of each reaction and variants with different resonance structures. After data augmentation, we have around 2 million entries to train the model, which achieves a testing set mean absolute error of 1.16 kcal mol-1 for the Gibbs free energy of activation in solution. We anticipate this model will accelerate predictions for high-throughput screening to quickly identify relevant reactions in solution, and our data set will serve as a benchmark for future studies.
- Top Cited Papers
- 6 August 2024
- Published by American Chemical Society (ACS) in Journal of the American Chemical Society
- Vol. 146 (33) , 23103-23120
When Do Quantum Mechanical Descriptors Help Graph Neural Networks to Predict Chemical Properties?
Deep graph neural networks are extensively utilized to predict chemical reactivity and molecular properties. However, because of the complexity of chemical space, such models often have difficulty extrapolating beyond the chemistry contained in the training set. Augmenting the model with quantum mechanical (QM) descriptors is anticipated to improve its generalizability. However, obtaining QM descriptors often requires CPU-intensive computational chemistry calculations. To identify when QM descriptors help graph neural networks predict chemical properties, we conduct a systematic investigation of the impact of atom, bond, and molecular QM descriptors on the performance of directed message passing neural networks (D-MPNNs) for predicting 16 molecular properties. The analysis surveys computational and experimental targets, as well as classification and regression tasks, and varied data set sizes from several hundred to hundreds of thousands of data points. Our results indicate that QM descriptors are mostly beneficial for D-MPNN performance on small data sets, provided that the descriptors correlate well with the targets and can be readily computed with high accuracy. Otherwise, using QM descriptors can add cost without benefit or even introduce unwanted noise that can degrade model performance. Strategic integration of QM descriptors with D-MPNN unlocks potential for physics-informed, data-efficient modeling with some interpretability that can streamline de novo drug and material designs. To facilitate the use of QM descriptors in machine learning workflows for chemistry, we provide a set of guidelines regarding when and how to best leverage QM descriptors, a high-throughput workflow to compute them, and an enhancement to Chemprop, a widely adopted open-source D-MPNN implementation for chemical property prediction.
- 6 August 2024
- Published by American Chemical Society (ACS) in Journal of the American Chemical Society
- Vol. 146 (33) , 23103-23120
Top Cited PapersWhen Do Quantum Mechanical Descriptors Help Graph Neural Networks to Predict Chemical Properties?
Deep graph neural networks are extensively utilized to predict chemical reactivity and molecular properties. However, because of the complexity of chemical space, such models often have difficulty extrapolating beyond the chemistry contained in the training set. Augmenting the model with quantum mechanical (QM) descriptors is anticipated to improve its generalizability. However, obtaining QM descriptors often requires CPU-intensive computational chemistry calculations. To identify when QM descriptors help graph neural networks predict chemical properties, we conduct a systematic investigation of the impact of atom, bond, and molecular QM descriptors on the performance of directed message passing neural networks (D-MPNNs) for predicting 16 molecular properties. The analysis surveys computational and experimental targets, as well as classification and regression tasks, and varied data set sizes from several hundred to hundreds of thousands of data points. Our results indicate that QM descriptors are mostly beneficial for D-MPNN performance on small data sets, provided that the descriptors correlate well with the targets and can be readily computed with high accuracy. Otherwise, using QM descriptors can add cost without benefit or even introduce unwanted noise that can degrade model performance. Strategic integration of QM descriptors with D-MPNN unlocks potential for physics-informed, data-efficient modeling with some interpretability that can streamline de novo drug and material designs. To facilitate the use of QM descriptors in machine learning workflows for chemistry, we provide a set of guidelines regarding when and how to best leverage QM descriptors, a high-throughput workflow to compute them, and an enhancement to Chemprop, a widely adopted open-source D-MPNN implementation for chemical property prediction.
- Open Access
- 1 July 2024
- Published by American Chemical Society (ACS) in ChemRxiv
When Do Quantum Mechanical Descriptors Help Graph Neural Networks Predict Chemical Properties?
Deep graph neural networks are extensively utilized to predict chemical reactivity and molecular properties. However, because of the complexity of chemical space, such models often have difficulty extrapolating beyond the chemistry contained in the training set. Augmented model with quantum mechanical (QM) descriptors is anticipated to improve its generalizability. However, obtaining QM descriptors often requires CPU-intensive computational chemistry calculations. To identify when QM descriptors help graph neural networks predict chemical properties, we conduct a systematic investigation of the impact of atom, bond, and molecular QM descriptors on the performance of directed message passing neural networks (D-MPNNs) for predicting 16 molecular properties. The analysis surveys computational and experimental targets, classification and regression tasks, and varied dataset sizes from several hundred to hundreds of thousands of datapoints. Our results indicate that QM descriptors are mostly beneficial for D-MPNN performance on small datasets, provided that the descriptors correlate well with the targets and can be readily computed with high accuracy. Otherwise, using QM descriptors can add cost without benefit or even introduce unwanted noise that can degrade model performance. Strategic integration of QM descriptors with D-MPNN unlocks potential for physics-informed, data-efficient modeling with some interpretability that can streamline de novo drug and material designs. To facilitate the use of QM descriptors in machine learning workflows for chemistry, we provide a set of guidelines regarding when and how to best leverage QM descriptors, a high-throughput workflow to compute them, and an enhancement to Chemprop, a widely adopted open-source D-MPNN implementation for chemical property prediction.
- 1 July 2024
- Published by American Chemical Society (ACS) in ChemRxiv
Open AccessWhen Do Quantum Mechanical Descriptors Help Graph Neural Networks Predict Chemical Properties?
Deep graph neural networks are extensively utilized to predict chemical reactivity and molecular properties. However, because of the complexity of chemical space, such models often have difficulty extrapolating beyond the chemistry contained in the training set. Augmented model with quantum mechanical (QM) descriptors is anticipated to improve its generalizability. However, obtaining QM descriptors often requires CPU-intensive computational chemistry calculations. To identify when QM descriptors help graph neural networks predict chemical properties, we conduct a systematic investigation of the impact of atom, bond, and molecular QM descriptors on the performance of directed message passing neural networks (D-MPNNs) for predicting 16 molecular properties. The analysis surveys computational and experimental targets, classification and regression tasks, and varied dataset sizes from several hundred to hundreds of thousands of datapoints. Our results indicate that QM descriptors are mostly beneficial for D-MPNN performance on small datasets, provided that the descriptors correlate well with the targets and can be readily computed with high accuracy. Otherwise, using QM descriptors can add cost without benefit or even introduce unwanted noise that can degrade model performance. Strategic integration of QM descriptors with D-MPNN unlocks potential for physics-informed, data-efficient modeling with some interpretability that can streamline de novo drug and material designs. To facilitate the use of QM descriptors in machine learning workflows for chemistry, we provide a set of guidelines regarding when and how to best leverage QM descriptors, a high-throughput workflow to compute them, and an enhancement to Chemprop, a widely adopted open-source D-MPNN implementation for chemical property prediction.
- 15 November 2023
- Published by American Chemical Society (ACS) in The Journal of Physical Chemistry B
- Vol. 127 (47) , 10151-10170
ConfSolv: Prediction of Solute Conformer-Free Energies across a Range of Solvents
Predicting Gibbs free energy of solution is key to understanding the solvent effects on thermodynamics and reaction rates for kinetic modeling. Accurately computing solution free energies requires the enumeration and evaluation of relevant solute conformers in solution. However, even after generation of relevant conformers, determining their free energy of solution requires an expensive workflow consisting of several ab initio computational chemistry calculations. To help address this challenge, we generate a large data set of solution free energies for nearly 44,000 solutes with almost 9 million conformers calculated in 41 different solvents using density functional theory and COSMO-RS and quantify the impact of solute conformers on the solution free energy. We then train a message passing neural network to predict the relative solution free energies of a set of solute conformers, enabling the identification of a small subset of thermodynamically relevant conformers. The model offers substantial computational time savings with predictions usually substantially within 1 kcal/mol of the free energy of the solution calculated by using computational chemical methods.
- 15 November 2023
- Published by American Chemical Society (ACS) in The Journal of Physical Chemistry B
- Vol. 127 (47) , 10151-10170
ConfSolv: Prediction of Solute Conformer-Free Energies across a Range of Solvents
Predicting Gibbs free energy of solution is key to understanding the solvent effects on thermodynamics and reaction rates for kinetic modeling. Accurately computing solution free energies requires the enumeration and evaluation of relevant solute conformers in solution. However, even after generation of relevant conformers, determining their free energy of solution requires an expensive workflow consisting of several ab initio computational chemistry calculations. To help address this challenge, we generate a large data set of solution free energies for nearly 44,000 solutes with almost 9 million conformers calculated in 41 different solvents using density functional theory and COSMO-RS and quantify the impact of solute conformers on the solution free energy. We then train a message passing neural network to predict the relative solution free energies of a set of solute conformers, enabling the identification of a small subset of thermodynamically relevant conformers. The model offers substantial computational time savings with predictions usually substantially within 1 kcal/mol of the free energy of the solution calculated by using computational chemical methods.
- Open Access
- 5 November 2023
- Published by The Open Journal in Journal of Open Source Software
- Vol. 8 (91) , 5996
Machine Learning Validation via Rational Dataset Sampling with astartes
Machine Learning (ML) has become an increasingly popular tool to accelerate traditional workflows. Critical to the use of ML is the process of splitting datasets into training, validation, and testing subsets that are used to develop and evaluate models. Common practice in the literature is to assign these subsets randomly. Although this approach is fast and efficient, it only measures a model's capacity to interpolate. Testing errors from random splits may be overly optimistic if given new data that is dissimilar to the scope of the training set; thus, there is a growing need to easily measure performance for extrapolation tasks. To address this issue, we report astartes, an open-source Python package that implements many similarityand distance-based algorithms to partition data into more challenging splits. Separate from astartes, users can then use these splits to better assess out-of-sample performance with any ML model of choice. This publication focuses on use-cases within cheminformatics. However, astartes operates on arbitrary vector inputs, so its principals and workflow are generalizable to other ML domains as well. astartes is available via the Python package managers pip and conda and is publicly hosted on GitHub (github.com/JacksonBurns/astartes).
- 5 November 2023
- Published by The Open Journal in Journal of Open Source Software
- Vol. 8 (91) , 5996
Open AccessMachine Learning Validation via Rational Dataset Sampling with astartes
Machine Learning (ML) has become an increasingly popular tool to accelerate traditional workflows. Critical to the use of ML is the process of splitting datasets into training, validation, and testing subsets that are used to develop and evaluate models. Common practice in the literature is to assign these subsets randomly. Although this approach is fast and efficient, it only measures a model's capacity to interpolate. Testing errors from random splits may be overly optimistic if given new data that is dissimilar to the scope of the training set; thus, there is a growing need to easily measure performance for extrapolation tasks. To address this issue, we report astartes, an open-source Python package that implements many similarityand distance-based algorithms to partition data into more challenging splits. Separate from astartes, users can then use these splits to better assess out-of-sample performance with any ML model of choice. This publication focuses on use-cases within cheminformatics. However, astartes operates on arbitrary vector inputs, so its principals and workflow are generalizable to other ML domains as well. astartes is available via the Python package managers pip and conda and is publicly hosted on GitHub (github.com/JacksonBurns/astartes).
- Open Access
- 6 October 2023
- Published by IOP Publishing in Machine Learning: Science and Technology
- Vol. 4 (4) , 048001
Comment on ‘Physics-based representations for machine learning properties of chemical reactions’
In a recent article in this journal, van Gerwen et al (2022 Mach. Learn.: Sci. Technol. 3 045005) presented a kernel ridge regression model to predict reaction barrier heights. Here, we comment on the utility of that model and present references and results that contradict several statements made in that article. Our primary interest is to offer a broader perspective by presenting three aspects that are essential for researchers to consider when creating models for chemical kinetics: (1) are the model’s prediction targets and associated errors sufficient for practical applications? (2) Does the model prioritize user-friendly inputs so it is practical for others to integrate into prediction workflows? (3) Does the analysis report performance on both interpolative and more challenging extrapolative data splits so users have a realistic idea of the likely errors in the model’s predictions?
- 6 October 2023
- Published by IOP Publishing in Machine Learning: Science and Technology
- Vol. 4 (4) , 048001
Open AccessComment on ‘Physics-based representations for machine learning properties of chemical reactions’
In a recent article in this journal, van Gerwen et al (2022 Mach. Learn.: Sci. Technol. 3 045005) presented a kernel ridge regression model to predict reaction barrier heights. Here, we comment on the utility of that model and present references and results that contradict several statements made in that article. Our primary interest is to offer a broader perspective by presenting three aspects that are essential for researchers to consider when creating models for chemical kinetics: (1) are the model’s prediction targets and associated errors sufficient for practical applications? (2) Does the model prioritize user-friendly inputs so it is practical for others to integrate into prediction workflows? (3) Does the analysis report performance on both interpolative and more challenging extrapolative data splits so users have a realistic idea of the likely errors in the model’s predictions?
- Open Access
- 3 April 2023
- Published by Wiley in International Journal of Chemical Kinetics
- Vol. 55 (6) , 300-323
Automated reaction kinetics and network exploration (Arkane): A statistical mechanics, thermodynamics, transition state theory, and master equation software
The open‐source statistical mechanics software described here, Arkane–Automated Reaction Kinetics and Network Exploration–facilitates computations of thermodynamic properties of chemical species, high‐pressure limit reaction rate coefficients, and pressure‐dependent rate coefficient over multi‐well molecular potential energy surfaces (PES) including the effects of collisional energy transfer on phenomenological kinetics. Arkane can use estimates to fill in information for molecules or reactions where quantum chemistry information is missing. The software solves the internal energy master equation for complex unimolecular reaction systems. Inputs to the software include converged electronic structure computations performed by the user using a variety of supported software packages (Gaussian, Molpro, Orca, TeraChem, Q‐Chem, Psi4). The software outputs high‐pressure limit rate coefficients and pressure‐dependent phenomenological rate coefficients, as well as computed thermodynamic properties (enthalpy, entropy, and constant pressure heat capacity) with added energy corrections. Some of the key features of Arkane include treatment of 1D, 2D or ND hindered internal rotation modes, treatment of free internal rotation modes, quantum tunneling effect consideration, transition state theory (TST) and Rice‐Ramsperger‐Kassel‐Marcus (RRKM) rate coefficient computations, master equation solution with four implemented methods, inverse‐Laplace transform of high‐pressure limit rate coefficients into the energy domain, energy corrections based on bond‐additivity or isodesmic reactions, automated and efficient PES exploration, and PES sensitivity analysis. The present work describes the design of Arkane, how it should be used, and refers to the theory that it employs. Arkane is distributed via the RMG‐Py software suite (https://github.com/ReactionMechanismGenerator/RMG‐Py).
- 3 April 2023
- Published by Wiley in International Journal of Chemical Kinetics
- Vol. 55 (6) , 300-323
Open AccessAutomated reaction kinetics and network exploration (Arkane): A statistical mechanics, thermodynamics, transition state theory, and master equation software
The open‐source statistical mechanics software described here, Arkane–Automated Reaction Kinetics and Network Exploration–facilitates computations of thermodynamic properties of chemical species, high‐pressure limit reaction rate coefficients, and pressure‐dependent rate coefficient over multi‐well molecular potential energy surfaces (PES) including the effects of collisional energy transfer on phenomenological kinetics. Arkane can use estimates to fill in information for molecules or reactions where quantum chemistry information is missing. The software solves the internal energy master equation for complex unimolecular reaction systems. Inputs to the software include converged electronic structure computations performed by the user using a variety of supported software packages (Gaussian, Molpro, Orca, TeraChem, Q‐Chem, Psi4). The software outputs high‐pressure limit rate coefficients and pressure‐dependent phenomenological rate coefficients, as well as computed thermodynamic properties (enthalpy, entropy, and constant pressure heat capacity) with added energy corrections. Some of the key features of Arkane include treatment of 1D, 2D or ND hindered internal rotation modes, treatment of free internal rotation modes, quantum tunneling effect consideration, transition state theory (TST) and Rice‐Ramsperger‐Kassel‐Marcus (RRKM) rate coefficient computations, master equation solution with four implemented methods, inverse‐Laplace transform of high‐pressure limit rate coefficients into the energy domain, energy corrections based on bond‐additivity or isodesmic reactions, automated and efficient PES exploration, and PES sensitivity analysis. The present work describes the design of Arkane, how it should be used, and refers to the theory that it employs. Arkane is distributed via the RMG‐Py software suite (https://github.com/ReactionMechanismGenerator/RMG‐Py).