Danish Khan
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Danish Khan has a total of 238 co-authors. Here are the top co-authors by number of shared publications:
- O. Anatole von Lilienfeld
Machine Learning Group Technische Universität Berlin 10587 Berlin Germany
6 shared publications - Adil Kabylda
University of Luxembourg
2 shared publications - Alastair J. A. Price
Department of Chemistry, Dalhousie University, 6274 Coburg Road, Dalhousie, Halifax, Canada
2 shared publications - Alexandre Tkatchenko
University of Luxembourg
2 shared publications - Anton Charkin-Gorbulin
University of Luxembourg
2 shared publications - Bing Huang
Wuhan University, Department of Chemistry and Molecular Sciences, 430072 Wuhan, China
2 shared publications
Latest Publications
- Open Access
- 22 June 2026
- Published by IOP Publishing in Electronic Structure
Roadmap on Advancements of the FHI-aims Software Package
Electronic-structure theory is the foundation of the description of materials including multiscale modeling of their properties and functions. Obviously, without sufficient accuracy at the base, reliable predictions are unlikely at any level that follows. The software package FHI-aims has proven to be a game changer for accurate free-energy calculations because of its scalability, numerical precision, and its efficient handling of density functional theory (DFT) with hybrid functionals and van der Waals interactions. It treats molecules, clusters, and extended systems (solids and liquids) on an equal footing. Besides DFT, FHI-aims also includes quantum-chemistry methods, descriptions for excited states and vibrations, and calculations of various types of transport. Recent advancements address the integration of FHI-aims into an increasing number of workflows and various artificial intelligence (AI) methods. This Roadmap describes the state-of-the-art of FHI-aims and advancements that are currently ongoing or planned.
- 22 June 2026
- Published by IOP Publishing in Electronic Structure
Open AccessRoadmap on Advancements of the FHI-aims Software Package
Electronic-structure theory is the foundation of the description of materials including multiscale modeling of their properties and functions. Obviously, without sufficient accuracy at the base, reliable predictions are unlikely at any level that follows. The software package FHI-aims has proven to be a game changer for accurate free-energy calculations because of its scalability, numerical precision, and its efficient handling of density functional theory (DFT) with hybrid functionals and van der Waals interactions. It treats molecules, clusters, and extended systems (solids and liquids) on an equal footing. Besides DFT, FHI-aims also includes quantum-chemistry methods, descriptions for excited states and vibrations, and calculations of various types of transport. Recent advancements address the integration of FHI-aims into an increasing number of workflows and various artificial intelligence (AI) methods. This Roadmap describes the state-of-the-art of FHI-aims and advancements that are currently ongoing or planned.
- 10 October 2025
- Published by AIP Publishing in The Journal of Chemical Physics
- Vol. 163 (14)
SCAN based non-linear double hybrid density functional
We develop a non-linear and non-empirical (nlane) double hybrid density functional derived from an accurate interpolation of the adiabatic connection in density functional theory, incorporating the correct asymptotic expansions. By bridging the second-order perturbative weak correlation limit with the fully interacting limit from the semi-local SCAN functional, nlane-SCAN is free of fitted parameters while providing improved energetic predictions compared to SCAN for moderately and strongly correlated systems alike. It delivers accurate predictions for atomic total energies and multiple reaction datasets from the GMTKN55 benchmark while significantly outperforming traditional linear hybrids and double hybrids for non-covalent interactions without requiring dispersion corrections. Due to the exact constraints at the weak correlation limit, nlane-SCAN has reduced delocalization errors, as evident through the SIE4x4 benchmark and bond dissociations of H2+ and He2+. Its proper asymptotic behavior ensures stability in strongly correlated systems, improving H2 and N2 bond dissociation profiles compared to conventional functionals.
- 10 October 2025
- Published by AIP Publishing in The Journal of Chemical Physics
- Vol. 163 (14)
SCAN based non-linear double hybrid density functional
We develop a non-linear and non-empirical (nlane) double hybrid density functional derived from an accurate interpolation of the adiabatic connection in density functional theory, incorporating the correct asymptotic expansions. By bridging the second-order perturbative weak correlation limit with the fully interacting limit from the semi-local SCAN functional, nlane-SCAN is free of fitted parameters while providing improved energetic predictions compared to SCAN for moderately and strongly correlated systems alike. It delivers accurate predictions for atomic total energies and multiple reaction datasets from the GMTKN55 benchmark while significantly outperforming traditional linear hybrids and double hybrids for non-covalent interactions without requiring dispersion corrections. Due to the exact constraints at the weak correlation limit, nlane-SCAN has reduced delocalization errors, as evident through the SIE4x4 benchmark and bond dissociations of H2+ and He2+. Its proper asymptotic behavior ensures stability in strongly correlated systems, improving H2 and N2 bond dissociation profiles compared to conventional functionals.
- Open Access
- 6 October 2025
- Published by National Academy of Sciences in Proceedings of the National Academy of Sciences
- Vol. 122 (41)
Generalized convolutional many-body distribution functional representations
Modern machine learning (ML) models of chemical and materials systems with billions of parameters require vast training datasets and considerable computational efforts. Lightweight kernel or decision tree-based methods, however, can be rapidly trained, leading to a considerably lower carbon footprint. We introduce generalized convolutional many-body distribution functionals (cMBDF) as highly compute and data-efficient atomic representations for accurate kernels that excel in low-data regimes. Generalizing the MBDF framework, cMBDF encodes local chemical environments in a compact fashion using translationally and rotationally invariant functionals of smooth atomic densities weighted by interaction potentials. The functional values can be efficiently evaluated by expressing them in terms of convolutions which are calculated via fast Fourier transforms and stored on predefined grids. In the generalized form, each atomic environment is described using a set of functionals uniformly defined by three integers; many-body, derivative, weighting orders. Irrespective of size/composition, cMBDF atomic vectors remain compact and constant in size for a fixed choice of these orders controlling the structural and compositional resolution. While being up to two orders of magnitude more compact than other popular representations, cMBDF is shown to be more accurate for the learning of various quantum properties such as energies, dipole moments, homo–lumo gaps, heat capacity, polarizability, optimal exact-exchange admixtures, and basis-set scaling factors. Applicability for organic and inorganic chemistry is tested as represented by the QM7b, QM9, and VQM24 datasets. Due to its compactness, model training and testing times are reduced from 23 h to 8 min, implying a corresponding reduction in carbon footprint.
- 6 October 2025
- Published by National Academy of Sciences in Proceedings of the National Academy of Sciences
- Vol. 122 (41)
Open AccessGeneralized convolutional many-body distribution functional representations
Modern machine learning (ML) models of chemical and materials systems with billions of parameters require vast training datasets and considerable computational efforts. Lightweight kernel or decision tree-based methods, however, can be rapidly trained, leading to a considerably lower carbon footprint. We introduce generalized convolutional many-body distribution functionals (cMBDF) as highly compute and data-efficient atomic representations for accurate kernels that excel in low-data regimes. Generalizing the MBDF framework, cMBDF encodes local chemical environments in a compact fashion using translationally and rotationally invariant functionals of smooth atomic densities weighted by interaction potentials. The functional values can be efficiently evaluated by expressing them in terms of convolutions which are calculated via fast Fourier transforms and stored on predefined grids. In the generalized form, each atomic environment is described using a set of functionals uniformly defined by three integers; many-body, derivative, weighting orders. Irrespective of size/composition, cMBDF atomic vectors remain compact and constant in size for a fixed choice of these orders controlling the structural and compositional resolution. While being up to two orders of magnitude more compact than other popular representations, cMBDF is shown to be more accurate for the learning of various quantum properties such as energies, dipole moments, homo–lumo gaps, heat capacity, polarizability, optimal exact-exchange admixtures, and basis-set scaling factors. Applicability for organic and inorganic chemistry is tested as represented by the QM7b, QM9, and VQM24 datasets. Due to its compactness, model training and testing times are reduced from 23 h to 8 min, implying a corresponding reduction in carbon footprint.
- Open Access
- 31 January 2025
- Published by American Association for the Advancement of Science (AAAS) in Science Advances
- Vol. 11 (5) , eadt7769
Adapting hybrid density functionals with machine learning
Exact exchange contributions significantly affect electronic states, influencing covalent bond formation and breaking. Hybrid density functional approximations, which average exact exchange admixtures empirically, have achieved success but fall short of high-level quantum chemistry accuracy due to delocalization errors. We propose adaptive hybrid functionals, generating optimal exact exchange admixture ratios on the fly using data-efficient quantum machine learning models with negligible overhead. The adaptive Perdew-Burke-Ernzerhof hybrid density functional (aPBE0) improves energetics, electron densities, and HOMO-LUMO gaps in QM9, QM7b, and GMTKN55 benchmark datasets. A model uncertainty-based constraint reduces the method smoothly to PBE0 in extrapolative regimes, ensuring general applicability with limited training. By tuning exact exchange fractions for different spin states, aPBE0 effectively addresses the spin gap problem in open-shell systems such as carbenes. We also present a revised QM9 (revQM9) dataset with more accurate quantum properties, including stronger covalent binding, larger bandgaps, more localized electron densities, and larger dipole moments.
- 31 January 2025
- Published by American Association for the Advancement of Science (AAAS) in Science Advances
- Vol. 11 (5) , eadt7769
Open AccessAdapting hybrid density functionals with machine learning
Exact exchange contributions significantly affect electronic states, influencing covalent bond formation and breaking. Hybrid density functional approximations, which average exact exchange admixtures empirically, have achieved success but fall short of high-level quantum chemistry accuracy due to delocalization errors. We propose adaptive hybrid functionals, generating optimal exact exchange admixture ratios on the fly using data-efficient quantum machine learning models with negligible overhead. The adaptive Perdew-Burke-Ernzerhof hybrid density functional (aPBE0) improves energetics, electron densities, and HOMO-LUMO gaps in QM9, QM7b, and GMTKN55 benchmark datasets. A model uncertainty-based constraint reduces the method smoothly to PBE0 in extrapolative regimes, ensuring general applicability with limited training. By tuning exact exchange fractions for different spin states, aPBE0 effectively addresses the spin gap problem in open-shell systems such as carbenes. We also present a revised QM9 (revQM9) dataset with more accurate quantum properties, including stronger covalent binding, larger bandgaps, more localized electron densities, and larger dipole moments.
- 22 January 2025
- Published by AIP Publishing in The Journal of Chemical Physics
- Vol. 162 (4)
Alchemical harmonic approximation based potential for iso-electronic diatomics: Foundational baseline for Δ-machine learning
We introduce the alchemical harmonic approximation (AHA) of the absolute electronic energy for charge-neutral iso-electronic diatomics at fixed interatomic distance d0. To account for variations in distance, we combine AHA with this ansatz for the electronic binding potential, E(d)=(Eu−Es)Ec−EsEu−Esd/d0+Es, where Eu, Ec, Es correspond to the energies of the united atom, calibration at d0, and the sum of infinitely separated atoms, respectively. Our model covers the two-dimensional electronic potential energy surface spanned by distances of 0.7–2.5 Å and differences in nuclear charge from which only one single point (with elements of nuclear charge Z1, Z2, and distance d0) is drawn to calibrate Ec. Using reference data from pbe0/cc-pVDZ, we present numerical evidence for the electronic ground-state of all neutral diatomics with 8, 10, 12, and 14 electrons. We assess the validity of our model by comparison to legacy interatomic potentials (harmonic oscillator, Lennard-Jones, and Morse) within the most relevant range of binding (0.7–2.5 Å) and find comparable accuracy if restricted to single diatomics and significantly better predictive power when extrapolating to the entire iso-electronic series. We also investigated Δ-learning of the electronic absolute energy using our model as a baseline. This baseline model results in a systematic improvement, effectively reducing training data needed for reaching chemical accuracy by up to an order of magnitude from ∼1000 to ∼100. By contrast, using AHA+Morse as a baseline hardly leads to any improvement and sometimes even deteriorates the predictive power. Inferring the energy of unseen CO converges to a prediction error of ∼0.1 Ha in direct learning and ∼0.04 Ha with our baseline.
- 22 January 2025
- Published by AIP Publishing in The Journal of Chemical Physics
- Vol. 162 (4)
Alchemical harmonic approximation based potential for iso-electronic diatomics: Foundational baseline for Δ-machine learning
We introduce the alchemical harmonic approximation (AHA) of the absolute electronic energy for charge-neutral iso-electronic diatomics at fixed interatomic distance d0. To account for variations in distance, we combine AHA with this ansatz for the electronic binding potential, E(d)=(Eu−Es)Ec−EsEu−Esd/d0+Es, where Eu, Ec, Es correspond to the energies of the united atom, calibration at d0, and the sum of infinitely separated atoms, respectively. Our model covers the two-dimensional electronic potential energy surface spanned by distances of 0.7–2.5 Å and differences in nuclear charge from which only one single point (with elements of nuclear charge Z1, Z2, and distance d0) is drawn to calibrate Ec. Using reference data from pbe0/cc-pVDZ, we present numerical evidence for the electronic ground-state of all neutral diatomics with 8, 10, 12, and 14 electrons. We assess the validity of our model by comparison to legacy interatomic potentials (harmonic oscillator, Lennard-Jones, and Morse) within the most relevant range of binding (0.7–2.5 Å) and find comparable accuracy if restricted to single diatomics and significantly better predictive power when extrapolating to the entire iso-electronic series. We also investigated Δ-learning of the electronic absolute energy using our model as a baseline. This baseline model results in a systematic improvement, effectively reducing training data needed for reaching chemical accuracy by up to an order of magnitude from ∼1000 to ∼100. By contrast, using AHA+Morse as a baseline hardly leads to any improvement and sometimes even deteriorates the predictive power. Inferring the energy of unseen CO converges to a prediction error of ∼0.1 Ha in direct learning and ∼0.04 Ha with our baseline.
- Open Access
- 27 September 2024
- Published by American Chemical Society (ACS) in ChemRxiv
Crash Testing Machine Learning Force Fields for Molecules, Materials, and Interfaces: Molecular Dynamics in the TEA Challenge 2023
We present the second part of the rigorous evaluation of state-of-the-art machine learning force fields (MLFFs) within the TEA Challenge 2023. This study provides an in-depth analysis of the performance of MACE, SO3krates, sGDML, SOAP/GAP, and FCHL19* in modeling molecules, molecule-surface interfaces, and periodic materials. We compare observables obtained from molecular dynamics (MD) simulations using different MLFFs under identical conditions. Where applicable, density-functional theory (DFT) or experiment serves as a reference to reliably assess the performance of the machine learning models. In the absence of DFT benchmarks, we conduct a comparative analysis based on results from various MLFF architectures. Our findings indicate that at the current stage of MLFF development, the choice of the ML model is becoming less critical, as the resulting simulations are largely independent of the MLFF architecture. Instead, emphasis should be placed on developing complete, reliable, and representative training datasets. Nonetheless, long-range noncovalent interactions remain challenging for all MLFF models, necessitating special caution in simulations of physical systems where such interactions are prominent, such as molecule-surface interfaces. The findings presented here reflect the state of MLFF models as of October 2023.
- 27 September 2024
- Published by American Chemical Society (ACS) in ChemRxiv
Open AccessCrash Testing Machine Learning Force Fields for Molecules, Materials, and Interfaces: Molecular Dynamics in the TEA Challenge 2023
We present the second part of the rigorous evaluation of state-of-the-art machine learning force fields (MLFFs) within the TEA Challenge 2023. This study provides an in-depth analysis of the performance of MACE, SO3krates, sGDML, SOAP/GAP, and FCHL19* in modeling molecules, molecule-surface interfaces, and periodic materials. We compare observables obtained from molecular dynamics (MD) simulations using different MLFFs under identical conditions. Where applicable, density-functional theory (DFT) or experiment serves as a reference to reliably assess the performance of the machine learning models. In the absence of DFT benchmarks, we conduct a comparative analysis based on results from various MLFF architectures. Our findings indicate that at the current stage of MLFF development, the choice of the ML model is becoming less critical, as the resulting simulations are largely independent of the MLFF architecture. Instead, emphasis should be placed on developing complete, reliable, and representative training datasets. Nonetheless, long-range noncovalent interactions remain challenging for all MLFF models, necessitating special caution in simulations of physical systems where such interactions are prominent, such as molecule-surface interfaces. The findings presented here reflect the state of MLFF models as of October 2023.
- Open Access
- 27 September 2024
- Published by American Chemical Society (ACS) in ChemRxiv
Crash Testing Machine Learning Force Fields for Molecules, Materials, and Interfaces: Model Analysis in the TEA Challenge 2023
Atomistic simulations are routinely employed in academia and industry to study the behavior of molecules, materials, and their interfaces. Central to these simulations are force fields (FF), whose development is challenged by intricate interatomic interactions at different spatio-temporal scales and the vast expanse of chemical space. Machine learning (ML) FFs, trained on quantum-mechanical energies and forces, have shown the capacity to achieve sub-kcal/(mol*A) accuracy while maintaining computational efficiency. The TEA Challenge 2023 rigorously evaluated commonly used MLFFs across diverse applications, highlighting their strengths and weaknesses. Participants trained their models using provided datasets, and the results were systematically analyzed to assess the MLFFs' ability to reproduce potential energy surfaces, handle incomplete reference data, manage multi-component systems, and model complex periodic structures. This publication describes the datasets, outlines the proposed challenges, and presents a detailed analysis of the accuracy, stability, and efficiency of the MACE, SO3krates, sGDML, SOAP/GAP, and FCHL19* architectures in molecular dynamics simulations. The models represent the MLFF developers who participated in the TEA Challenge 2023. All results presented correspond to the state of the ML architectures as of October 2023. A comprehensive analysis of the molecular dynamics results obtained with different MLFFs will be presented in the second part of this manuscript.
- 27 September 2024
- Published by American Chemical Society (ACS) in ChemRxiv
Open AccessCrash Testing Machine Learning Force Fields for Molecules, Materials, and Interfaces: Model Analysis in the TEA Challenge 2023
Atomistic simulations are routinely employed in academia and industry to study the behavior of molecules, materials, and their interfaces. Central to these simulations are force fields (FF), whose development is challenged by intricate interatomic interactions at different spatio-temporal scales and the vast expanse of chemical space. Machine learning (ML) FFs, trained on quantum-mechanical energies and forces, have shown the capacity to achieve sub-kcal/(mol*A) accuracy while maintaining computational efficiency. The TEA Challenge 2023 rigorously evaluated commonly used MLFFs across diverse applications, highlighting their strengths and weaknesses. Participants trained their models using provided datasets, and the results were systematically analyzed to assess the MLFFs' ability to reproduce potential energy surfaces, handle incomplete reference data, manage multi-component systems, and model complex periodic structures. This publication describes the datasets, outlines the proposed challenges, and presents a detailed analysis of the accuracy, stability, and efficiency of the MACE, SO3krates, sGDML, SOAP/GAP, and FCHL19* architectures in molecular dynamics simulations. The models represent the MLFF developers who participated in the TEA Challenge 2023. All results presented correspond to the state of the ML architectures as of October 2023. A comprehensive analysis of the molecular dynamics results obtained with different MLFFs will be presented in the second part of this manuscript.
- Open Access
- 18 July 2023
- Published by AIP Publishing in The Journal of Chemical Physics
- Vol. 159 (3)
Kernel based quantum machine learning at record rate: Many-body distribution functionals as compact representations
The feature vector mapping used to represent chemical systems is a key factor governing the superior data efficiency of kernel based quantum machine learning (QML) models applicable throughout chemical compound space. Unfortunately, the most accurate representations require a high dimensional feature mapping, thereby imposing a considerable computational burden on model training and use. We introduce compact yet accurate, linear scaling QML representations based on atomic Gaussian many-body distribution functionals (MBDF) and their derivatives. Weighted density functions of MBDF values are used as global representations that are constant in size, i.e., invariant with respect to the number of atoms. We report predictive performance and training data efficiency that is competitive with state-of-the-art for two diverse datasets of organic molecules, QM9 and QMugs. Generalization capability has been investigated for atomization energies, highest occupied molecular orbital–lowest unoccupied molecular orbital eigenvalues and gap, internal energies at 0 K, zero point vibrational energies, dipole moment norm, static isotropic polarizability, and heat capacity as encoded in QM9. MBDF based QM9 performance lowers the optimal Pareto front spanned between sampling and training cost to compute node minutes, effectively sampling chemical compound space with chemical accuracy at a sampling rate of ∼48 molecules per core second.
- 18 July 2023
- Published by AIP Publishing in The Journal of Chemical Physics
- Vol. 159 (3)
Open AccessKernel based quantum machine learning at record rate: Many-body distribution functionals as compact representations
The feature vector mapping used to represent chemical systems is a key factor governing the superior data efficiency of kernel based quantum machine learning (QML) models applicable throughout chemical compound space. Unfortunately, the most accurate representations require a high dimensional feature mapping, thereby imposing a considerable computational burden on model training and use. We introduce compact yet accurate, linear scaling QML representations based on atomic Gaussian many-body distribution functionals (MBDF) and their derivatives. Weighted density functions of MBDF values are used as global representations that are constant in size, i.e., invariant with respect to the number of atoms. We report predictive performance and training data efficiency that is competitive with state-of-the-art for two diverse datasets of organic molecules, QM9 and QMugs. Generalization capability has been investigated for atomization energies, highest occupied molecular orbital–lowest unoccupied molecular orbital eigenvalues and gap, internal energies at 0 K, zero point vibrational energies, dipole moment norm, static isotropic polarizability, and heat capacity as encoded in QM9. MBDF based QM9 performance lowers the optimal Pareto front spanned between sampling and training cost to compute node minutes, effectively sampling chemical compound space with chemical accuracy at a sampling rate of ∼48 molecules per core second.
- Open Access
An Interacting Quantum Atoms (IQA) and Relative Energy Gradient (REG) Analysis of the Anomeric Effect
The explanation of the anomeric effect in terms of underlying quantum properties is still controversial almost 70 years after its introduction. Here, we use a method called Relative Energy Gradient (REG), which is able to compute chemical insight with a view to explaining the anomeric effect. REG operates on atomic energy contributions generated by the quantum topological energy decomposition Interacting Quantum Atoms (IQA). Based on the case studies of dimethoxymethane and 2-fluorotetrahydropyran, we show that the anomeric effect is electrostatic in nature rather than governed by hyperconjugation.
Open AccessAn Interacting Quantum Atoms (IQA) and Relative Energy Gradient (REG) Analysis of the Anomeric Effect
The explanation of the anomeric effect in terms of underlying quantum properties is still controversial almost 70 years after its introduction. Here, we use a method called Relative Energy Gradient (REG), which is able to compute chemical insight with a view to explaining the anomeric effect. REG operates on atomic energy contributions generated by the quantum topological energy decomposition Interacting Quantum Atoms (IQA). Based on the case studies of dimethoxymethane and 2-fluorotetrahydropyran, we show that the anomeric effect is electrostatic in nature rather than governed by hyperconjugation.
- 10 June 2022
- Published by Elsevier in Journal of Molecular Graphics and Modelling
- Vol. 116, 108251
Mechanism of phenol and p-nitrophenol adsorption on kaolinite surface in aqueous medium: A molecular dynamics study
- 10 June 2022
- Published by Elsevier in Journal of Molecular Graphics and Modelling
- Vol. 116, 108251
Mechanism of phenol and p-nitrophenol adsorption on kaolinite surface in aqueous medium: A molecular dynamics study