Structural Descriptors Research Articles

Insight to the prediction of CO2 solubility in ionic liquids based on the interpretable machine learning model

In this work, we investigated three different machine learning (ML)-based models, i.e., gaussian process regression (GPR), LightGBM, and CatBoost, for predicting the solubility of CO2 in various ionic liquids (ILs). Three molecular descriptors, i.e., group contribution (GC), molecular structure descriptors (MSD), and hybrid GC-MSD are used in our three models. The performance of our developed models were rigorously evaluated using mean absolute error (MAE), coefficient of determination (R2), and mean relative error (MRE) (i.e., relative deviation in percentage), with each model subjected to multiple tests employing different random state parameters. The dataset underwent partitioning into training and testing sets at an 80:20 ratio, with additional splits at various ratios to assess prediction performance sensitivity. Overall, all models exhibited proficient CO2 solubility prediction in ILs, with performance varying based on descriptor type. Notably, the hybrid GC-MSD consistently outperformed others, attributed to GC-MSD incorporates a broader array of molecular feature information. Particularly, the CatBoost-GC-MSD model excelled, achieving an impressive R2 of 0.9925, MAE of 0.0122, and MRE of 11.1550%. Comparing our models to previous studies revealed the superior performance of CatBoost-GC-MSD across all descriptor types. Furthermore, our model interpretation, employing shapley additive explanation (SHAP) analysis, identified pressure, temperature, Chi0, Kappa2, and EState_VSA10 as the top five influential input features. These findings provide valuable insights into the underlying molecular features affecting CO2 solubility in ILs and lay the foundation for future research in this field.

Characterizing grain boundary network length features through a harmonic representation

When modeling and characterizing grain boundary networks (GBNs), there are situations where local descriptors such as triple junction fractions (TJFs) and special boundary fractions are identical between two microstructures, but the performance or properties between the two are distinct. These differences are caused by higher length-scale features that cannot be identified using local structural descriptors. Spectral graph theory (SGT) has been used previously to encode network length features, but did not enable direct quantitative comparisons between the structures that cause property differences. In this paper, we derive a harmonic representation of diffusion on GBNs based on SGT. This method enables direct quantitative comparisons between GBNs for microstructures with different morphologies, and identifies network length features responsible for structural and performance differences, which local descriptors cannot explain. We show an interpretation of the eigenmodes generated by this method that explains long-range structural causes of certain property differences. We apply this method to a large library of microstructures, and identify structural classes through clustering. We show that equal proportioned TJF and J1 dominated microstructures are the most sensitive to network length differences because of boundary configurations, while J2-J3 dominated structures are the least sensitive. This method also identifies network length features that result in anomalous percolation/non-percolation compared to predictions based on local correlations alone.

Customized Virtual File System

Customized Virtual File System

Development of a Robust Read-Across Model for the Prediction of Biological Potency of Novel Peroxisome Proliferator-Activated Receptor Delta Agonists.

A robust predictive model was developed using 136 novel peroxisome proliferator-activated receptor delta (PPARδ) agonists, a distinct subtype of lipid-activated transcription factors of the nuclear receptor superfamily that regulate target genes by binding to characteristic sequences of DNA bases. The model employs various structural descriptors and docking calculations and provides predictions of the biological activity of PPARδ agonists, following the criteria of the Organization for Economic Co-operation and Development (OECD) for the development and validation of quantitative structure-activity relationship (QSAR) models. Specifically focused on small molecules, the model facilitates the identification of highly potent and selective PPARδ agonists and offers a read-across concept by providing the chemical neighbours of the compound under study. The model development process was conducted on Isalos Analytics Software (v. 0.1.17) which provides an intuitive environment for machine-learning applications. The final model was released as a user-friendly web tool and can be accessed through the Enalos Cloud platform's graphical user interface (GUI).

Machine Learning‐Enhanced Prediction of Inorganic Semiconductor Bandgaps for Advancing Optoelectronic Technologies

AbstractA pivotal challenge in advancing inorganics optoelectronic technologies, is the precise characterization of materials' electronic attributes, with the bandgap being a critical property. Conventional approaches, heavily reliant on time‐intensive and financially demanding experimental and computational methods, such as density functional theory (DFT) calculations, face limitations due to inherent estimation errors. Machine learning methodologies are developed for the prediction of bandgaps of inorganic semiconductors but most of them are employed for datasets created by DFT calculations, hence limiting their performance. Addressing this, the study leverages machine learning methodologies, harnessing both compositional and structural features, to predict the band gaps of inorganic semiconductors with enhanced accuracy. This advancement is reinforced by the employment of an experimental bandgap dataset, which, when integrated with structural descriptors obtained from the Materials Project, significantly improves prediction capabilities. This is evidenced by the model's exceptional performance across two distinct benchmark datasets. Furthermore, the model's adeptness in predicting formation energies underscores its versatility and applicability to a broad spectrum of electronic properties. These findings suggest that the predictive accuracy of this model can be further augmented through the inclusion of additional experimental bandgap measurements and the refinement of structural descriptors. This approach offers a promising and efficient alternative to traditional methodologies, potentially accelerating the development of optoelectronic technologies.

Quantitative read-across structure-activity relationship (q-RASAR): A novel approach to estimate the subchronic oral safety (NOAEL) of diverse organic chemicals in rats

We have developed a quantitative safety prediction model for subchronic repeated doses of diverse organic chemicals on rats using the novel quantitative read-across structure-activity relationship (q-RASAR) approach, which uses similarity-based descriptors for predictive model generation. The experimental -Log (NOAEL) values have been used here as a potential indicator of oral subchronic safety on rats as it determines the maximum dose level for which no observed adverse effects of chemicals are found. A total of 186 data points of diverse organic chemicals have been used for the model generation using structural and physicochemical (0D–2D) descriptors. The read-across-derived similarity, error, and concordance measures (RASAR descriptors) have been extracted from the preliminary 0D–2D descriptors. Then, the combined pool of RASAR and the identified 0D–2D descriptors of the training set were employed to develop the final models by using the partial least squares (PLS) algorithm. The developed PLS model was rigorously validated by various internal and external validation metrics as suggested by the Organization for Economic Co-operation and Development (OECD). The final q-RASAR model is proven to be statistically sound, robust and externally predictive (R2 = 0.85, Q2LOO = 0.82 and Q2F1 = 0.94), superseding the internal as well as external predictivity of the corresponding quantitative structure-activity relationship (QSAR) model as well as previously reported subchronic repeated dose toxicity model found in the literature. In a nutshell, the q-RASAR is an effective approach that has the potential to be used as a good alternative way to improve external predictivity, interpretability, and transferability for subchronic oral safety prediction as well as ecotoxicity risk identification.

Synthesis and structural characterization of picolinamide and thiol chelated Pt(II) complexes: Their kinetics & mechanism and theoretical approach

Picolinamide is a versatile chelating ligand and it reveals peculiar behaviour for metal chelation, particularly for its amide group. The bonding mode of N & O donor centres of an amide group depends on the metal ion, their associated charge and the other associated ligands in metal complexes. Picolinamide (pica) based Pt(II) complex C1 to C4 were synthesized and their bonding mode with Pt(II) was established by different spectroscopic analyses and theoretical approaches for their structural importance and potential anticancer activity. The complex [Pt(pica)Cl2]; C1 was synthesized and its hydrolyzed product [Pt(pica)(OH2)2](ClO4)2; C2 was obtained from hydrolysis of complex C1. The kinetic and mechanistic study for the interaction of complex C2 with l-cysteine (l-cys) and N-acetyl-l-cysteine (N-ac-l-cys) was performed in pseudo-first-order reaction condition to evaluate the rate constants, rate equation, enthalpy of activation (ΔHǂ) and entropy of activation (ΔSǂ). Sulfur chelated complex [Pt(pica)(l-cys)]+; C3 and [Pt(pica)(N-ac-l-cys)]; C4 were obtained when a 1:1 M ratio of C2 and thiol are mixed separately. Theoretical investigation on drug likeness property was evaluated by PASS prediction software. The theoretical approach on structural properties like; structural optimization, HOMO-LUMO energy, TD-DFT simulation, MEP and reactivity descriptors of the complexes were executed and analyzed with the Gaussian09 software package. Molecular docking of the complexes was executed with DNA and BSA.

The Euler Sombor index of a graph

AbstractA novel geometric method is proposed for constructing vertex‐degree‐based molecular structure descriptors, which is the perimeter of an ellipse whose focal points represent the degree‐point and its dual‐point of a pair of adjacent vertices in a graph . We first analyze the chemical applicability of this elliptic Sombor index, and then establish its mathematical properties and determine the extremal values for this elliptic Sombor index among all (molecular) trees and characterize the corresponding extremal graphs.

RIGA: Rotation-Invariant and Globally-Aware Descriptors for Point Cloud Registration.

Successful point cloud registration relies on accurate correspondences established upon powerful descriptors. However, existing neural descriptors either leverage a rotation-variant backbone whose performance declines under large rotations, or encode local geometry that is less distinctive. To address this issue, we introduce RIGA to learn descriptors that are Rotation-Invariant by design and Globally-Aware. From the Point Pair Features (PPFs) of sparse local regions, rotation-invariant local geometry is encoded into geometric descriptors. Global awareness of 3D structures and geometric context is subsequently incorporated, both in a rotation-invariant fashion. More specifically, 3D structures of the whole frame are first represented by our global PPF signatures, from which structural descriptors are learned to help geometric descriptors sense the 3D world beyond local regions. Geometric context from the whole scene is then globally aggregated into descriptors. Finally, the description of sparse regions is interpolated to dense point descriptors, from which correspondences are extracted for registration. To validate our approach, we conduct extensive experiments on both object- and scene-level data. With large rotations, RIGA surpasses the state-of-the-art methods by a margin of 8 ° in terms of the Relative Rotation Error on ModelNet40 and improves the Feature Matching Recall by at least 5 percentage points on 3DLoMatch.

Posidonia oceanica (L.) Delile, 1813 seedlings as a potential integrative tool for restoring damaged meadows

Seagrasses provide key ecological services in coastal marine and estuarine environments. Human activities and environmental changes affect seagrasses compromising their structure and functioning. In this context, seagrass restoration has become a worldwide priority to halt and reverse degradation and to recover ecosystem functionality and associated services. Here, we investigate the sexual reproduction of Posidonia oceanica by describing the flowering and fruiting process. Furthermore, we compare the structural (survival and shoot density) and functional (daily leaf production, DLP) descriptors of seedlings with those derived from a P. oceanica transplantation performed through cuttings to assess their feasibility as an integrative source of vegetal material for restoration purposes. Despite many initial flowers and fruits, a small number of mature fruits are produced, confirming the low reproductive success of the phanerogam. A comparison of structural and functional features showed that seedlings and cuttings showed similar survival and shoot density rates after more than 2 years. In addition, DLP showed that seedlings increased in leaf length over the study period, reaching a similar value registered by cuttings. The study demonstrates the effectiveness of using seedlings in restoration efforts. It also emphasizes the importance of exploring P. oceanica reproductive ecology to better understand temporal and spatial variations in flowering, as well as fruiting success, at local and basin scales.

Using atomic clustering based on structural and electronic descriptors that consider surrounding environment to evaluate local properties of DFT functionals.

We developed a method for evaluating the accuracies of the local properties of DFT functionals in detail using a clustering method based on machine learning and structural/electronic descriptors. We generated 36 clusters consistent with human intuition using 30,436 carbon atoms from the QM9 dataset. The results were used to evaluate 13C NMR chemical shifts calculated using 84 DFT functionals. Carbon atoms were grouped based on their similar environments, reducing errors within these groups. This enables more accurate assessment of the accuracy using a specific DFT functional. Therefore, the present atomic clustering provides more detailed insight into accuracy verification.

Exploring Layered Disorder in Lithium-Ion-Conducting Li3Y1-xInxCl6.

Li3Y1-xInxCl6 undergoes a phase transition from trigonal to monoclinic via an intermediate orthorhombic phase. Although the trigonal yttrium containing the end member phase, Li3YCl6, synthesized by a mechanochemical route, is known to exhibit stacking fault disorder, not much is known about the monoclinic phases of the serial composition Li3Y1-xInxCl6. This work aims to shed light on the influence of the indium substitution on the phase evolution, along with the evolution of stacking fault disorder using X-ray and neutron powder diffraction together with solid-state nuclear magnetic resonance spectroscopy, studying the lithium-ion diffusion. Although Li3Y1-xInxCl6 with x ≤ 0.1 exhibits an ordered trigonal structure like Li3YCl6, a large degree of stacking fault disorder is observed in the monoclinic phases for the x ≥ 0.3 compositions. The stacking fault disorder materializes as a crystallographic intergrowth of faultless domains with staggered layers stacked in a uniform layer stacking, along with faulted domains with randomized staggered layer stacking. This work shows how structurally complex even the "simple" series of solid solutions can be in this class of halide-based lithium-ion conductors, as apparent from difficulties in finding a consistent structural descriptor for the ionic transport.

Machine Learning-Based Prediction of Reduction Potentials for PtIV Complexes.

Some of the well-known drawbacks of clinically approved PtII complexes can be overcome using six-coordinate PtIV complexes as inert prodrugs, which release the corresponding four-coordinate active PtII species upon reduction by cellular reducing agents. Therefore, the key factor of PtIV prodrug mechanism of action is their tendency to be reduced which, when the involved mechanism is of outer-sphere type, is measured by the value of the reduction potential. Machine learning (ML) models can be used to effectively capture intricate relationships within PtIV complex data, leading to highly accurate predictions of reduction potentials and other properties, and offering significant insights into their electrochemical behavior and potential applications. In this study, a machine learning-based approach for predicting the reduction potentials of PtIV complexes based on relevant molecular descriptors is presented. Leveraging a data set of experimentally determined reduction potentials and a diverse range of molecular descriptors, the proposed model demonstrates remarkable predictive accuracy (MSE = 0.016 V2, RMSE = 0.13 V, R2 = 0.92). Ab initio calculations and a set of different machine learning algorithms and feature engineering techniques have been employed to systematically explore the relationship between molecular structure and similarity and reduction potential. Specifically, it has been investigated whether the reduction potential of these compounds can be described by combining ML models across different combinations of constitutional, topological, and electronic molecular descriptors. Our results not only provide insights into the crucial factors influencing reduction potentials but also offer a rapid and effective tool for the rational design of PtIV complexes with tailored electrochemical properties for pharmaceutical applications. This approach has the potential to significantly expedite the development and screening of novel PtIV prodrug candidates. The analysis of principal components and key features extracted from the model highlights the significance of structural descriptors of the 2D Atom Pairs type and the lowest unoccupied molecular orbital energy. Specifically, with just 20 appropriately selected descriptors, a notable separation of complexes based on their reduction potential value is achieved.

Interpretable Machine Learning Models for Practical Antimonate Electrocatalyst Performance.

Computationally predicting the performance of catalysts under reaction conditions is a challenging task due to the complexity of catalytic surfaces and their evolution in situ, different reaction paths, and the presence of solid-liquid interfaces in the case of electrochemistry. We demonstrate here how relatively simple machine learning models can be found that enable prediction of experimentally observed onset potentials. Inputs to our model are comprised of data from the oxygen reduction reaction on non-precious transition-metal antimony oxide nanoparticulate catalysts with a combination of experimental conditions and computationally affordable bulk atomic and electronic structural descriptors from density functional theory simulations. From human-interpretable genetic programming models, we identify key experimental descriptors and key supplemental bulk electronic and atomic structural descriptors that govern trends in onset potentials for these oxides and deduce how these descriptors should be tuned to increase onset potentials. We finally validate these machine learning predictions by experimentally confirming that scandium as a dopant in nickel antimony oxide leads to a desired onset potential increase. Macroscopic experimental factors are found to be crucially important descriptors to be considered for models of catalytic performance, highlighting the important role machine learning can play here even in the presence of small datasets.

A magnetic/force coupling assisted lithium-oxygen battery based on magnetostriction and piezoelectric catalysis of CoFe2O4/BiFeO3 cathode

The high energy density of Li−O2 batteries surpasses all existing batteries, and it holds the potential to emerge as the most outstanding solution for energy storage in the future. However, the insulated, insoluble discharge product (Li2O2) has impeded the practical applications. Conventional catalyst design based on electronic structure and interfacial charge transfer descriptors are incapable of overcoming these limitations of Li2O2. Herein, a magnetic/force coupling assisted Li−O2 battery based on magnetostrictive and piezoelectric catalysis with CoFe2O4/BiFeO3 core-shell structure cathode was established for the first time. An external magnetic field is introduced to produce a magnetostrictive stress on CoFe2O4, contributing to the piezoelectric electron hole transport by the generated piezoelectric potential energy with a built-in electric field based on the piezoelectric catalytic mechanism, and thus promoting the oxygen reduction reaction (ORR) and oxygen evolution reaction (OER) kinetics, and reducing the overpotential during charge/discharge. The magnetic/force coupling assisted Li−O2 battery equipped with the unique CoFe2O4/BiFeO3 cathode deliver an ultra-low charging platform of 3.49 V and an ultra-high discharge platform of 2.83 V. This unique magnetic/force assisted strategy provides an important insight into solving the high overpotential in metal-air battery systems.

Prediction of Bond Dissociation Energy for Organic Molecules Based on a Machine‐Learning Approach

Comprehensive SummaryBond dissociation energy (BDE), which refers to the enthalpy change for the homolysis of a specific covalent bond, is one of the basic thermodynamic properties of molecules. It is very important for understanding chemical reactivities, chemical properties and chemical transformations. Here, a machine learning‐based comprehensive BDE prediction model was established based on the iBonD experimental BDE dataset and the calculated BDE dataset by St. John et al. Differential Structural and PhysicOChemical (D‐SPOC) descriptors that reflected changes in molecules’ structural and physicochemical features in the process of bond homolysis were designed as input features. The model trained with LightGBM algorithm gave a low mean absolute error (MAE) of 1.03 kcal/mol on the test set. The D‐SPOC model could apply to accurate BDE prediction of phenol O—H bonds, uncommon N‐SCF3 and O‐SCF3 reagents, and β‐C—H bonds in enamine intermediates. A fast online prediction platform was constructed based on the D‐SPOC model, which could be found at http://isyn.luoszgroup.com/bde_prediction.

Understanding Defects in Amorphous Silicon with Million-Atom Simulations and Machine Learning.

The structure of amorphous silicon (a-Si) is widely thought of as a fourfold-connected random network, and yet it is defective atoms, with fewer or more than four bonds, that make it particularly interesting. Despite many attempts to explain such "dangling-bond" and "floating-bond" defects, respectively, a unified understanding is still missing. Here, we use advanced computational chemistry methods to reveal the complex structural and energetic landscape of defects in a-Si. We study an ultra-large-scale, quantum-accurate structural model containing a million atoms, and thousands of individual defects, allowing reliable defect-related statistics to be obtained. We combine structural descriptors and machine-learned atomic energies to develop a classification of the different types of defects in a-Si. The results suggest a revision of the established floating-bond model by showing that fivefold-bonded atoms in a-Si exhibit a wide range of local environments-analogous to fivefold centers in coordination chemistry. Furthermore, it is shown that fivefold (but not threefold) coordination defects tend to cluster together. Our study provides new insights into one of the most widely studied amorphous solids, and has general implications for understanding defects in disordered materials beyond silicon alone.

Understanding Defects in Amorphous Silicon with Million‐Atom Simulations and Machine Learning

AbstractThe structure of amorphous silicon (a‐Si) is widely thought of as a fourfold‐connected random network, and yet it is defective atoms, with fewer or more than four bonds, that make it particularly interesting. Despite many attempts to explain such “dangling‐bond” and “floating‐bond” defects, respectively, a unified understanding is still missing. Here, we use advanced computational chemistry methods to reveal the complex structural and energetic landscape of defects in a‐Si. We study an ultra‐large‐scale, quantum‐accurate structural model containing a million atoms, and thousands of individual defects, allowing reliable defect‐related statistics to be obtained. We combine structural descriptors and machine‐learned atomic energies to develop a classification of the different types of defects in a‐Si. The results suggest a revision of the established floating‐bond model by showing that fivefold‐bonded atoms in a‐Si exhibit a wide range of local environments–analogous to fivefold centers in coordination chemistry. Furthermore, it is shown that fivefold (but not threefold) coordination defects tend to cluster together. Our study provides new insights into one of the most widely studied amorphous solids, and has general implications for understanding defects in disordered materials beyond silicon alone.

Chemometrics-driven prediction and prioritization of diverse pesticides on chickens for addressing hazardous effects on public health

The extensive use of various pesticides in the agriculture field badly affects both chickens and humans, primarily through residues in food products and environmental exposure. This study offers the first quantitative structure-toxicity relationship (QSTR) and quantitative read-across-structure toxicity relationship (q-RASTR) models encompassing the LOEL and NOEL endpoints for acute toxicity in chicken, a widely consumed protein. The study's significance lies in the direct link between chemical toxicity in chicken, human intake, and environmental damage. Both the QSTR and the similarity-based read-across algorithms are applied concurrently to improve the predictability of the models. The q-RASTR models were generated by combining read-across derived similarity and error-based parameters, alongside structural and physicochemical descriptors. Machine Learning approaches (SVM and RR) were also employed with the optimization of relevant hyperparameters based on the cross-validation approach, and the final test set prediction results were compared. The PLS-based q-RASTR models for NOEL and LOEL endpoints showed good statistical performance, as traced from the external validation metrics Q2F1: 0.762–0.844; Q2F2: 0.759–0.831 and MAEtest: 0.195–0.214. The developed models were further used to screen the Pesticide Properties DataBase (PPDB) for potential toxicants in chickens. Thus, established models can address eco-toxicological data gaps and development of novel and safe eco-friendly pesticides.

The rule of four: anomalous distributions in the stoichiometries of inorganic compounds

Why are materials with specific characteristics more abundant than others? This is a fundamental question in materials science and one that is traditionally difficult to tackle, given the vastness of compositional and configurational space. We highlight here the anomalous abundance of inorganic compounds whose primitive unit cell contains a number of atoms that is a multiple of four. This occurrence—named here the rule of four—has to our knowledge not previously been reported or studied. Here, we first highlight the rule’s existence, especially notable when restricting oneself to experimentally known compounds, and explore its possible relationship with established descriptors of crystal structures, from symmetries to energies. We then investigate this relative abundance by looking at structural descriptors, both of global (packing configurations) and local (the smooth overlap of atomic positions) nature. Contrary to intuition, the overabundance does not correlate with low-energy or high-symmetry structures; in fact, structures which obey the rule of four are characterized by low symmetries and loosely packed arrangements maximizing the free volume. We are able to correlate this abundance with local structural symmetries, and visualize the results using a hybrid supervised-unsupervised machine learning method.