The development of deep learning has made non-biochemical methods for molecular property prediction screening a reality,which can increase the experimental speed and reduce the experimental cost of relevant experiment...The development of deep learning has made non-biochemical methods for molecular property prediction screening a reality,which can increase the experimental speed and reduce the experimental cost of relevant experiments.There are currently two main approaches to representing molecules:(a)representing molecules by fixing molecular descriptors,and(b)representing molecules by graph convolutional neural networks.Currently,both of these Representative methods have achieved some results in their respective experiments.Based on past efforts,we propose a Dual Self-attention Fusion Message Neural Network(DSFMNN).DSFMNN uses a combination of dual self-attention mechanism and graph convolutional neural network.Advantages of DSFMNN:(1)The dual self-attention mechanism focuses not only on the relationship between individual subunits in a molecule but also on the relationship between the atoms and chemical bonds contained in each subunit.(2)On the directed molecular graph,a message delivery approach centered on directed molecular bonds is used.We test the performance of the model on eight publicly available datasets and compare the performance with several models.Based on the current experimental results,DSFMNN has superior performance compared to previous models on the datasets applied in this paper.展开更多
Accurate prediction of molecular properties is crucial for selecting compounds with ideal properties and reducing the costs and risks of trials.Traditional methods based on manually crafted features and graph-based me...Accurate prediction of molecular properties is crucial for selecting compounds with ideal properties and reducing the costs and risks of trials.Traditional methods based on manually crafted features and graph-based methods have shown promising results in molecular property prediction.However,traditional methods rely on expert knowledge and often fail to capture the complex structures and interactions within molecules.Similarly,graph-based methods typically overlook the chemical structure and function hidden in molecular motifs and struggle to effectively integrate global and local molecular information.To address these limitations,we propose a novel fingerprint-enhanced hierarchical graph neural network(FH-GNN)for molecular property prediction that simultaneously learns information from hierarchical molecular graphs and fingerprints.The FH-GNN captures diverse hierarchical chemical information by applying directed message-passing neural networks(D-MPNN)on a hierarchical molecular graph that integrates atomic-level,motif-level,and graph-level information along with their relationships.Addi-tionally,we used an adaptive attention mechanism to balance the importance of hierarchical graphs and fingerprint features,creating a comprehensive molecular embedding that integrated hierarchical mo-lecular structures with domain knowledge.Experiments on eight benchmark datasets from MoleculeNet showed that FH-GNN outperformed the baseline models in both classification and regression tasks for molecular property prediction,validating its capability to comprehensively capture molecular informa-tion.By integrating molecular structure and chemical knowledge,FH-GNN provides a powerful tool for the accurate prediction of molecular properties and aids in the discovery of potential drug candidates.展开更多
1 Introduction Molecular Property Prediction aims to identify molecules sharing similar efficacious properties[1],which is a foundational task in drug discovery,materials science and bioinformatics.Graph neural networ...1 Introduction Molecular Property Prediction aims to identify molecules sharing similar efficacious properties[1],which is a foundational task in drug discovery,materials science and bioinformatics.Graph neural networks(GNNs)have shown significant success in this field.However,GNN-based methods often face label scarcity,limiting their performance in predicting molecular properties.Besides,GNNs trained on specific datasets frequently struggle with generalization due to domain shift[2].展开更多
Chemical industry is always seeking opportunities to efficiently and economically convert raw materials to commodity chemicals and higher value-added chemicalbased products.The life cycles of chemical products involve...Chemical industry is always seeking opportunities to efficiently and economically convert raw materials to commodity chemicals and higher value-added chemicalbased products.The life cycles of chemical products involve the procedures of conceptual product designs,experimental investigations,sustainable manufactures through appropriate chemical processes and waste disposals.During these periods,one of the most important keys is the molecular property prediction models associating molecular structures with product properties.In this paper,a framework combining quantum mechanics and quantitative structure-property relationship is established for fast molecular property predictions,such as activity coefficient,and so forth.The workflow of framework consists of three steps.In the first step,a database is created for collections of basic molecular information;in the second step,quantum mechanics-based calculations are performed to predict quantum mechanics-based/derived molecular properties(pseudo experimental data),which are stored in a database and further provided for the developments of quantitative structure-property relationship methods for fast predictions of properties in the third step.The whole framework has been carried out within a molecular property prediction toolbox.Two case studies highlighting different aspects of the toolbox involving the predictions of heats of reaction and solid-liquid phase equilibriums are presented.展开更多
The prediction of molecular properties is a crucial task in the field of drug discovery.Computational methods that can accurately predict molecular properties can significantly accelerate the drug discovery process an...The prediction of molecular properties is a crucial task in the field of drug discovery.Computational methods that can accurately predict molecular properties can significantly accelerate the drug discovery process and reduce the cost of drug discovery.In recent years,iterative updates in computing hardware and the rise of deep learning have created a new and effective path for molecular property prediction.Deep learning methods can leverage the vast amount of data accumulated over the years in drug discovery and do not require complex feature engineering.in this review,we summarize molecular representations and commonly used datasets in molecular property prediction models and present advanced deep learning methods for molecular property prediction,including state-of-the-art deep learning networks such as graph neural networks and Transformer-based models,as well as state-of-the-art deep learning strategies such as 3D pre-train,contrastive learning,multi-task learning,transfer learning,and meta-learning.We also point out some critical issues such as lack of datasets,low information utilization,and lack of specificity for diseases.展开更多
The precise prediction of molecular properties is essential for advancements in drug development,particularly in virtual screening and compound optimization.The recent introduction of numerous deep learningbased metho...The precise prediction of molecular properties is essential for advancements in drug development,particularly in virtual screening and compound optimization.The recent introduction of numerous deep learningbased methods has shown remarkable potential in enhancing Molecular Property Prediction(MPP),especially improving accuracy and insights into molecular structures.Yet,two critical questions arise:does the integration of domain knowledge augment the accuracy of molecular property prediction and does employing multi-modal data fusion yield more precise results than unique data source methods?To explore these matters,we comprehensively review and quantitatively analyze recent deep learning methods based on various benchmarks.We discover that integrating molecular information significantly improves Molecular Property Prediction(MPP)for both regression and classification tasks.Specifically,regression improvements,measured by reductions in Root Mean Square Error(RMSE),are up to 4.0%,while classification enhancements,measured by the area under the receiver operating characteristic curve(ROC-AUC),are up to 1.7%.Additionally,we discover that,as measured by ROC-AUC,augmenting 2D graphs with 3D information improves performance for classification tasks by up to 13.2%and enriching 2D graphs with 1D SMILES boosts multi-modal learning performance for regression tasks by up to 9.1%.The two consolidated insights offer crucial guidance for future advancements in drug discovery.展开更多
The modulation of electrical properties of MoS_2 has attracted extensive research interest because of its potential applications in electronic and optoelectronic devices.Herein,interfacial charge transfer induced elec...The modulation of electrical properties of MoS_2 has attracted extensive research interest because of its potential applications in electronic and optoelectronic devices.Herein,interfacial charge transfer induced electronic property tuning of MoS_2 are investigated by in situ ultraviolet photoelectron spectroscopy and x-ray photoelectron spectroscopy measurements.A downward band-bending of MoS_2-related electronic states along with the decreasing work function,which are induced by the electron transfer from Cs overlayers to MoS_2,is observed after the functionalization of MoS_2 with Cs,leading to n-type doping.Meanwhile,when MoS_2 is modified with 2,3,5,6-tetrafluoro-7,7,8,8-tetracyanoquinodimethane(F_4-TCNQ),an upward band-bending of MoS_2-related electronic states along with the increasing work function is observed at the interfaces.This is attributed to the electron depletion within MoS_2 due to the strong electron withdrawing property of F_4-TCNQ,indicating p-type doping of MoS_2.Our findings reveal that surface transfer doping is an effective approach for electronic property tuning of MoS_2 and paves the way to optimize its performance in electronic and optoelectronic devices.展开更多
Dissolved organic matter(DOM)represents the largest pool of reactive carbon on the Earth and plays a crucial role in various biogeochemical processes and ecosystem functions.However,it is understudied for a global und...Dissolved organic matter(DOM)represents the largest pool of reactive carbon on the Earth and plays a crucial role in various biogeochemical processes and ecosystem functions.However,it is understudied for a global understanding of DOM molecular properties such as molecular weight,stoichiometry,and oxidation state,and the linkages among them across Earth systems.Here,a meta-analysis of 2707 sites in 204 literatures was conducted by synthesizing four representative molecular properties of DOM,i.e.,mass,double bond equivalent(DBE),modified aromaticity index(AI_(mod)),and nominal oxidation state of carbon(NOSC).By exploring H/C and O/C ratios,we examined the relationships among these DOM properties across waters and land systems,and their geographical patterns and environmental drivers.We found that,compared to land system,the mass,DBE,and AI_(mod) were all significantly higher in water systems,with river sediments exhibiting the highest values.The DOM oxidation state indicated by NOSC was greater on average in wastewater(NOSC=0.226±0.06)and marine water(NOSC=0.133±0.06)than in other habitats.Compared to waters,the mass in land system showed more strongly positive correlations with oxidation states such as NOSC and O/C,and the NOSC showed stronger relations to bioavailability properties such as DBE,AI_(mod),and H/C.Among all the properties,H/C and AI_(mod) contributed to the most variations in global DOM properties.In waters,NOSC monotonically increased towards high latitudes,while DBE and AI_(mod) showed significant hump-shaped patterns indicating peaked unsaturation and aromaticity at mid-latitudes of approximately absolute 30°–50°.The variations in DOM properties were significantly correlated with environmental factors such as annual mean temperature and pH.Collectively,we revealed the spatial distribution and environmental drivers of DOM molecular properties across Earth ecosystems,which could shed light on our comprehensive understanding of DOM characteristics and its dynamics.展开更多
9,10-Phenanthrenequinone（PQ） and benzil are important a-diketones. This manuscript explains the first comparison of PQ and benzil molecular properties. We have used 1H NMR, 13C NMR, 1H-IH COSY, HMBC, HMQC, UV-Vis ab...9,10-Phenanthrenequinone（PQ） and benzil are important a-diketones. This manuscript explains the first comparison of PQ and benzil molecular properties. We have used 1H NMR, 13C NMR, 1H-IH COSY, HMBC, HMQC, UV-Vis absorption and emission, CV and TGA experiments to study PQ and benzil that provided the following novel results. （1） The 1H NMR（CDC13） of PQ show δ 8.19（H1）, 8.02（H4）, 7.72（H3）, 7.47（H2） instead of an earlier reported 8.25（H4）, 8.08（H1）, 7.80（H2）, 7.55（H3）; （2） in the 13C NMR（CDCl3）, the C9/C10（C=O） signal of PQ appears upfield（6 180.3） compared to C9/Cl0（C=O） signal of benzil（6 194.5）, which shows higher electrophilic character（more attractive for nucleophiles） of C9/C10（C=O） of benzil; （3） the first 2max for the UV-Vis absorption and emission of PQ are blue-shifted compared to benzil despite increased conjugation attributed to the different symmetries（C2v for PQ and C2h for Benzil） of the two molecules; （4） the emission spectrum of benzil is broader compared to that of PQ due to slower relaxation of the excited state; （5） The CV study shows that PQ and benzil are good electron acceptors and PQ shows a better reduction process than benzil due to an extra ring that provides stability for the reduced species（mono or diradical anions）; （6） TGA shows the higher thermal stability of PQ than benzil attributed to the presence of phenanthrene unit in PQ.展开更多
In this paper,an ab initio,local density functional(LDF)method was used to explore the relationship between the molecular properties of additives and the lubricating performance of aluminum rolling oil.The structural ...In this paper,an ab initio,local density functional(LDF)method was used to explore the relationship between the molecular properties of additives and the lubricating performance of aluminum rolling oil.The structural properties of butyl stearate,dodecanol,docosanol,and methyl dodecanoate were studied according to the density functional theory.The calculated data showed that the atoms in or around the functional groups might be likely the reacting sites.Because of the different functional groups and structure of ester and alcohol,two types of complex additives,dodecanol and butyl stearate,methyl dodecanoate and butyl stearate,respectively,were chosen for studying their tribological properties and performing aluminum cold rolling experiments.The test results agreed with the calculated results very well.The complex ester,viz.methyl dodecanoate and butyl stearate,had the best lubricating performance with a friction coefficient of 0.084 1 and a permissive-rolling thickness of 0.040 mm as compared with that of dodecanol-butyl stearate-base oil formulation.展开更多
Flower blight on anthurium(Anthurium andraeanum)was observed during August 2018 on an anthurium cultivation farm in the Songkhla Province of southern Thailand.The fungal isolate was identified as Neopestalotiopsis cla...Flower blight on anthurium(Anthurium andraeanum)was observed during August 2018 on an anthurium cultivation farm in the Songkhla Province of southern Thailand.The fungal isolate was identified as Neopestalotiopsis clavispora based on the morphology and DNA sequence of the internal transcribed spacer(ITS),translation elongation factor 1-α(tef1-α),andβ-tubulin(tub)genes.The phylogenetic tree,based on the combined sequences of ITS,tef1-α,and tub,confirmed this pathogen as N.clavispora.Pathogenicity of the species was confirmed according to Koch’s postulate:N.clavispora could infect anthurium.To the best of our knowledge,this is the first report of N.clavispora as a pathogen of anthurium.展开更多
This work focuses on the relationship between flexibility of molecular chains and thermal properties of polyurethane elastomer（PUE）, which laid the foundation of further research about how to improve thermal propert...This work focuses on the relationship between flexibility of molecular chains and thermal properties of polyurethane elastomer（PUE）, which laid the foundation of further research about how to improve thermal properties of PUE. A series of PUE samples with different flexibility of molecular chains was prepared by using 1,4-butanediol（1,4-BDO）/bisphenol-a（BPA） blends with different mole ratios including9/1, 8/2, 7/3, 6/4 and 5/5. As comparison, PUE extended with pure 1,4-BDO and BPA was also synthesized.These samples were characterized by differential scanning calorimetry（DSC）, thermogravimetric analysis（TGA）, dynamic mechanical analysis（DMA）, etc. The results showed that with the decrease in flexibility of molecular chains the glass transition temperature（Tg） increased and low-temperature properties became worse. Besides, all samples had a certain degree of microphase separation, and soft segments in some samples were crystallized, i.e. the decreasing flexibility of molecular chains led to the impossibility of chains tightly packing and crystalline domains forming so that the degree of microphase separation decreased and the thermal properties became worse.展开更多
In this paper, a new theory for the dynamic properties of protein molecular system have been proposed by means of the analysis of the characteristics of the collective excitations generated under the localized fluctua...In this paper, a new theory for the dynamic properties of protein molecular system have been proposed by means of the analysis of the characteristics of the collective excitations generated under the localized fluctuation and the deformation of structure of the protein. Some new results obtained from this study show that the Davydov theory is an approximate and disadvantegeous theory.展开更多
Based on the nonequilibrium Green function method and density functional theory calculations, we theoretically investigate the effect of chirality on the electronic transport properties of thioxanthene-based molecular...Based on the nonequilibrium Green function method and density functional theory calculations, we theoretically investigate the effect of chirality on the electronic transport properties of thioxanthene-based molecular switch. The molecule comprises the switch which can exhibit different chiralities, that is, cis-form and trans-form by ultraviolet or visible irradiation. The results clearly reveal that the switching behaviors can be realized when the molecule converts between cis-form and trans-form. ~urthermore, the on-off ratio can be modulated by the chirality of the carbon nanotube electrodes. The maximum on-off ratio can reach 109 at 0.4 V for the armchair junction, suggesting potential applications of this type of junctions in future design of functional molecular devices.展开更多
Based on molecular mechanics and the deformation characteristics of the atomic lattice structure of graphene, a modifi ed molecular structure mechanics method was developed to improve the original one, that is, the se...Based on molecular mechanics and the deformation characteristics of the atomic lattice structure of graphene, a modifi ed molecular structure mechanics method was developed to improve the original one, that is, the semi-rigid connections were used to model the bond angle variations between the C-Cbonds in graphene. The simulated results show that the equivalent space frame model with semi-rigid connections for graphene proposed in this article is a simple, efficient, and accurate model to evaluate the equivalent elastic properties of graphene. Though the present computational model of the semi-rigid connected space frame is only applied to characterize the mechanical behaviors of the space lattices of graphene, it has more potential applications in the static and dynamic analyses of graphene and other nanomaterials.展开更多
We focus on two new 21） materials, i.e., monolayer and bilayer silicon phosphides （Sil P1）. Based on the elastic- scattering Green＇s function, the electronic-transport properties of two-dimensional monolayer and b...We focus on two new 21） materials, i.e., monolayer and bilayer silicon phosphides （Sil P1）. Based on the elastic- scattering Green＇s function, the electronic-transport properties of two-dimensional monolayer and bilayer Au- Si1P1-Au molecular junctions are studied. It is found that their bandgaps are narrow （0.16eV for a monolayer molecular junction and 0.26 e V for a bilayer molecular junction）. Moreover, the calculated current-voltage char- acteristics indicate that the monolayer molecular junction provides constant output current （20 hA） over a wide voltage range, and the bilayer molecular junction provides higher current （42 hA）.展开更多
We investigate the electronic-transport properties of two-dimensional monolayer films from Au-P-Au molecular junction to Au-Si-Au molecular junction using elastic scattering Green＇s function theory. In the process of...We investigate the electronic-transport properties of two-dimensional monolayer films from Au-P-Au molecular junction to Au-Si-Au molecular junction using elastic scattering Green＇s function theory. In the process of replacing the P atoms with Si atoms every other line from the middle of monolayer blue phosphorus molecular structure, the substitution of Si atoms changes the properties of Au-P-Au molecular junction significantly. Interestingly, the current value has a symmetric change as a parabolic curve with the peak appearing in Au-Si_1P_1-Au molecular junction, which provides the most stable current of 15.00 nA in a wide voltage range of 0.70-2.70 V.Moreover, the current-voltage characteristics of the structures indicate that the steps tend to disappear revealing the property similar to metal when the Si atoms dominate the molecular junction.展开更多
Through molecular dynamics（MD） simulation, the dependencies of temperature, grain size and strain rate on the mechanical properties were studied. The simulation results demonstrated that the strain rate from 0.05 to...Through molecular dynamics（MD） simulation, the dependencies of temperature, grain size and strain rate on the mechanical properties were studied. The simulation results demonstrated that the strain rate from 0.05 to 2 ns–1 affected the Young＇s modulus of nickel nanowires slightly, whereas the yield stress increased. The Young＇s modulus decreased approximately linearly; however, the yield stress firstly increased and subsequently dropped as the temperature increased. The Young＇s modulus and yield stress increased as the mean grain size increased from 2.66 to 6.72 nm. Moreover, certain efforts have been made in the microstructure evolution with mechanical properties association under uniaxial tension. Certain phenomena such as the formation of twin structures, which were found in nanowires with larger grain size at higher strain rate and lower temperature, as well as the movement of grain boundaries and dislocation, were detected and discussed in detail. The results demonstrated that the plastic deformation was mainly accommodated by the motion of grain boundaries for smaller grain size. However, for larger grain size, the formations of stacking faults and twins were the main mechanisms of plastic deformation in the polycrystalline nickel nanowire.展开更多
Finding energetic materials with tailored properties is always a significant challenge due to low research efficiency in trial and error.Herein,a methodology combining domain knowledge,a machine learning algorithm,and...Finding energetic materials with tailored properties is always a significant challenge due to low research efficiency in trial and error.Herein,a methodology combining domain knowledge,a machine learning algorithm,and experiments is presented for accelerating the discovery of novel energetic materials.A high-throughput virtual screening(HTVS)system integrating on-demand molecular generation and machine learning models covering the prediction of molecular properties and crystal packing mode scoring is established.With the proposed HTVS system,candidate molecules with promising properties and a desirable crystal packing mode are rapidly targeted from the generated molecular space containing 25112 molecules.Furthermore,a study of the crystal structure and properties shows that the good comprehensive performances of the target molecule are in agreement with the predicted results,thus verifying the effectiveness of the proposed methodology.This work demonstrates a new research paradigm for discovering novel energetic materials and can be extended to other organic materials without manifest obstacles.展开更多
文摘The development of deep learning has made non-biochemical methods for molecular property prediction screening a reality,which can increase the experimental speed and reduce the experimental cost of relevant experiments.There are currently two main approaches to representing molecules:(a)representing molecules by fixing molecular descriptors,and(b)representing molecules by graph convolutional neural networks.Currently,both of these Representative methods have achieved some results in their respective experiments.Based on past efforts,we propose a Dual Self-attention Fusion Message Neural Network(DSFMNN).DSFMNN uses a combination of dual self-attention mechanism and graph convolutional neural network.Advantages of DSFMNN:(1)The dual self-attention mechanism focuses not only on the relationship between individual subunits in a molecule but also on the relationship between the atoms and chemical bonds contained in each subunit.(2)On the directed molecular graph,a message delivery approach centered on directed molecular bonds is used.We test the performance of the model on eight publicly available datasets and compare the performance with several models.Based on the current experimental results,DSFMNN has superior performance compared to previous models on the datasets applied in this paper.
基金supported by Macao Science and Technology Development Fund,Macao SAR,China(Grant No.:0043/2023/AFJ)the National Natural Science Foundation of China(Grant No.:22173038)Macao Polytechnic University,Macao SAR,China(Grant No.:RP/FCA-01/2022).
文摘Accurate prediction of molecular properties is crucial for selecting compounds with ideal properties and reducing the costs and risks of trials.Traditional methods based on manually crafted features and graph-based methods have shown promising results in molecular property prediction.However,traditional methods rely on expert knowledge and often fail to capture the complex structures and interactions within molecules.Similarly,graph-based methods typically overlook the chemical structure and function hidden in molecular motifs and struggle to effectively integrate global and local molecular information.To address these limitations,we propose a novel fingerprint-enhanced hierarchical graph neural network(FH-GNN)for molecular property prediction that simultaneously learns information from hierarchical molecular graphs and fingerprints.The FH-GNN captures diverse hierarchical chemical information by applying directed message-passing neural networks(D-MPNN)on a hierarchical molecular graph that integrates atomic-level,motif-level,and graph-level information along with their relationships.Addi-tionally,we used an adaptive attention mechanism to balance the importance of hierarchical graphs and fingerprint features,creating a comprehensive molecular embedding that integrated hierarchical mo-lecular structures with domain knowledge.Experiments on eight benchmark datasets from MoleculeNet showed that FH-GNN outperformed the baseline models in both classification and regression tasks for molecular property prediction,validating its capability to comprehensively capture molecular informa-tion.By integrating molecular structure and chemical knowledge,FH-GNN provides a powerful tool for the accurate prediction of molecular properties and aids in the discovery of potential drug candidates.
基金sponsored in part by the National Key Research and Development Program of China(No.2023YFB3307500)the Science and Technology Innovation Project of Hunan Province(No.2023RC4014)the National Natural Science Foundation of China(NSFC)(Grant Nos.62076146,62021002,U20A6003,6212780016).
文摘1 Introduction Molecular Property Prediction aims to identify molecules sharing similar efficacious properties[1],which is a foundational task in drug discovery,materials science and bioinformatics.Graph neural networks(GNNs)have shown significant success in this field.However,GNN-based methods often face label scarcity,limiting their performance in predicting molecular properties.Besides,GNNs trained on specific datasets frequently struggle with generalization due to domain shift[2].
基金The authors are grateful for the financial supports of the National Natural Science Foundation of China(Grant Nos.22078041 and 21808025)the Fundamental Research Funds for the Central Universities(Grant No.DUT20JC41).
文摘Chemical industry is always seeking opportunities to efficiently and economically convert raw materials to commodity chemicals and higher value-added chemicalbased products.The life cycles of chemical products involve the procedures of conceptual product designs,experimental investigations,sustainable manufactures through appropriate chemical processes and waste disposals.During these periods,one of the most important keys is the molecular property prediction models associating molecular structures with product properties.In this paper,a framework combining quantum mechanics and quantitative structure-property relationship is established for fast molecular property predictions,such as activity coefficient,and so forth.The workflow of framework consists of three steps.In the first step,a database is created for collections of basic molecular information;in the second step,quantum mechanics-based calculations are performed to predict quantum mechanics-based/derived molecular properties(pseudo experimental data),which are stored in a database and further provided for the developments of quantitative structure-property relationship methods for fast predictions of properties in the third step.The whole framework has been carried out within a molecular property prediction toolbox.Two case studies highlighting different aspects of the toolbox involving the predictions of heats of reaction and solid-liquid phase equilibriums are presented.
基金National Natural Science Foundation of China,Grant/Award Number:62071278Macao PolytechnicUniversity,Grant/Award Numbers:RP/FCSD-01/2022,RP/FCSD-02/2022。
文摘The prediction of molecular properties is a crucial task in the field of drug discovery.Computational methods that can accurately predict molecular properties can significantly accelerate the drug discovery process and reduce the cost of drug discovery.In recent years,iterative updates in computing hardware and the rise of deep learning have created a new and effective path for molecular property prediction.Deep learning methods can leverage the vast amount of data accumulated over the years in drug discovery and do not require complex feature engineering.in this review,we summarize molecular representations and commonly used datasets in molecular property prediction models and present advanced deep learning methods for molecular property prediction,including state-of-the-art deep learning networks such as graph neural networks and Transformer-based models,as well as state-of-the-art deep learning strategies such as 3D pre-train,contrastive learning,multi-task learning,transfer learning,and meta-learning.We also point out some critical issues such as lack of datasets,low information utilization,and lack of specificity for diseases.
文摘The precise prediction of molecular properties is essential for advancements in drug development,particularly in virtual screening and compound optimization.The recent introduction of numerous deep learningbased methods has shown remarkable potential in enhancing Molecular Property Prediction(MPP),especially improving accuracy and insights into molecular structures.Yet,two critical questions arise:does the integration of domain knowledge augment the accuracy of molecular property prediction and does employing multi-modal data fusion yield more precise results than unique data source methods?To explore these matters,we comprehensively review and quantitatively analyze recent deep learning methods based on various benchmarks.We discover that integrating molecular information significantly improves Molecular Property Prediction(MPP)for both regression and classification tasks.Specifically,regression improvements,measured by reductions in Root Mean Square Error(RMSE),are up to 4.0%,while classification enhancements,measured by the area under the receiver operating characteristic curve(ROC-AUC),are up to 1.7%.Additionally,we discover that,as measured by ROC-AUC,augmenting 2D graphs with 3D information improves performance for classification tasks by up to 13.2%and enriching 2D graphs with 1D SMILES boosts multi-modal learning performance for regression tasks by up to 9.1%.The two consolidated insights offer crucial guidance for future advancements in drug discovery.
基金Supported by the National Natural Science Foundation of China (Grant No.22002031)the Natural Science Foundation of Zhejiang Province (Grant No.LY18F010019)the Innovation Project in Hangzhou for Returned Scholar。
文摘The modulation of electrical properties of MoS_2 has attracted extensive research interest because of its potential applications in electronic and optoelectronic devices.Herein,interfacial charge transfer induced electronic property tuning of MoS_2 are investigated by in situ ultraviolet photoelectron spectroscopy and x-ray photoelectron spectroscopy measurements.A downward band-bending of MoS_2-related electronic states along with the decreasing work function,which are induced by the electron transfer from Cs overlayers to MoS_2,is observed after the functionalization of MoS_2 with Cs,leading to n-type doping.Meanwhile,when MoS_2 is modified with 2,3,5,6-tetrafluoro-7,7,8,8-tetracyanoquinodimethane(F_4-TCNQ),an upward band-bending of MoS_2-related electronic states along with the increasing work function is observed at the interfaces.This is attributed to the electron depletion within MoS_2 due to the strong electron withdrawing property of F_4-TCNQ,indicating p-type doping of MoS_2.Our findings reveal that surface transfer doping is an effective approach for electronic property tuning of MoS_2 and paves the way to optimize its performance in electronic and optoelectronic devices.
基金supported by the National Natural Science Foundation of China(Nos.U24A20578,42225708,42377122,92251304)the Basic Research Program of Jiangsu Province(No.BK20240111)+1 种基金the Key Laboratory of Lake and Watershed Science for Water Security(No.NKL2023-QN04)the Science and Technology Planning Project of Nanjing Institute of Geography and Limnology,Chinese Academy of Sciences(No.NIGLAS2022GS09).
文摘Dissolved organic matter(DOM)represents the largest pool of reactive carbon on the Earth and plays a crucial role in various biogeochemical processes and ecosystem functions.However,it is understudied for a global understanding of DOM molecular properties such as molecular weight,stoichiometry,and oxidation state,and the linkages among them across Earth systems.Here,a meta-analysis of 2707 sites in 204 literatures was conducted by synthesizing four representative molecular properties of DOM,i.e.,mass,double bond equivalent(DBE),modified aromaticity index(AI_(mod)),and nominal oxidation state of carbon(NOSC).By exploring H/C and O/C ratios,we examined the relationships among these DOM properties across waters and land systems,and their geographical patterns and environmental drivers.We found that,compared to land system,the mass,DBE,and AI_(mod) were all significantly higher in water systems,with river sediments exhibiting the highest values.The DOM oxidation state indicated by NOSC was greater on average in wastewater(NOSC=0.226±0.06)and marine water(NOSC=0.133±0.06)than in other habitats.Compared to waters,the mass in land system showed more strongly positive correlations with oxidation states such as NOSC and O/C,and the NOSC showed stronger relations to bioavailability properties such as DBE,AI_(mod),and H/C.Among all the properties,H/C and AI_(mod) contributed to the most variations in global DOM properties.In waters,NOSC monotonically increased towards high latitudes,while DBE and AI_(mod) showed significant hump-shaped patterns indicating peaked unsaturation and aromaticity at mid-latitudes of approximately absolute 30°–50°.The variations in DOM properties were significantly correlated with environmental factors such as annual mean temperature and pH.Collectively,we revealed the spatial distribution and environmental drivers of DOM molecular properties across Earth ecosystems,which could shed light on our comprehensive understanding of DOM characteristics and its dynamics.
基金Supported by National Natural Science Foundation of China(Nos.20704016,20573040,20474024,20125421,90101026,50303007)Ministry of Education(No.20070183202)
文摘9,10-Phenanthrenequinone（PQ） and benzil are important a-diketones. This manuscript explains the first comparison of PQ and benzil molecular properties. We have used 1H NMR, 13C NMR, 1H-IH COSY, HMBC, HMQC, UV-Vis absorption and emission, CV and TGA experiments to study PQ and benzil that provided the following novel results. （1） The 1H NMR（CDC13） of PQ show δ 8.19（H1）, 8.02（H4）, 7.72（H3）, 7.47（H2） instead of an earlier reported 8.25（H4）, 8.08（H1）, 7.80（H2）, 7.55（H3）; （2） in the 13C NMR（CDCl3）, the C9/C10（C=O） signal of PQ appears upfield（6 180.3） compared to C9/Cl0（C=O） signal of benzil（6 194.5）, which shows higher electrophilic character（more attractive for nucleophiles） of C9/C10（C=O） of benzil; （3） the first 2max for the UV-Vis absorption and emission of PQ are blue-shifted compared to benzil despite increased conjugation attributed to the different symmetries（C2v for PQ and C2h for Benzil） of the two molecules; （4） the emission spectrum of benzil is broader compared to that of PQ due to slower relaxation of the excited state; （5） The CV study shows that PQ and benzil are good electron acceptors and PQ shows a better reduction process than benzil due to an extra ring that provides stability for the reduced species（mono or diradical anions）; （6） TGA shows the higher thermal stability of PQ than benzil attributed to the presence of phenanthrene unit in PQ.
基金the financial support of this study provided by the National Natural Science Foundation of China(No.51274037)the Cooperation Program between USTB and SINOPEC(No.112116)
文摘In this paper,an ab initio,local density functional(LDF)method was used to explore the relationship between the molecular properties of additives and the lubricating performance of aluminum rolling oil.The structural properties of butyl stearate,dodecanol,docosanol,and methyl dodecanoate were studied according to the density functional theory.The calculated data showed that the atoms in or around the functional groups might be likely the reacting sites.Because of the different functional groups and structure of ester and alcohol,two types of complex additives,dodecanol and butyl stearate,methyl dodecanoate and butyl stearate,respectively,were chosen for studying their tribological properties and performing aluminum cold rolling experiments.The test results agreed with the calculated results very well.The complex ester,viz.methyl dodecanoate and butyl stearate,had the best lubricating performance with a friction coefficient of 0.084 1 and a permissive-rolling thickness of 0.040 mm as compared with that of dodecanol-butyl stearate-base oil formulation.
基金supported by Prince of Songkla Universitythe Center of Excellence in Agricultural and Natural Resources Biotechnology(Grant No.CoE-ANRB)phase 3。
文摘Flower blight on anthurium(Anthurium andraeanum)was observed during August 2018 on an anthurium cultivation farm in the Songkhla Province of southern Thailand.The fungal isolate was identified as Neopestalotiopsis clavispora based on the morphology and DNA sequence of the internal transcribed spacer(ITS),translation elongation factor 1-α(tef1-α),andβ-tubulin(tub)genes.The phylogenetic tree,based on the combined sequences of ITS,tef1-α,and tub,confirmed this pathogen as N.clavispora.Pathogenicity of the species was confirmed according to Koch’s postulate:N.clavispora could infect anthurium.To the best of our knowledge,this is the first report of N.clavispora as a pathogen of anthurium.
基金supported financially by the National Natural Science Foundation of China (Grant No. 51372200)Program for New Century Excellent Talents in University of Ministry of Education of China (Grant No. NCET-12-1045)+2 种基金Special Program for local serving from Education Department of Shaanxi Provincial Government (Grant No. 2013JC19)Program for Innovation Team in Xi’an University of Technology (Grant No. 108-25605T401)Ph.D. Innovation Fund Projects of Xi’an University of Technology (Fund No. 310-252071501)
文摘This work focuses on the relationship between flexibility of molecular chains and thermal properties of polyurethane elastomer（PUE）, which laid the foundation of further research about how to improve thermal properties of PUE. A series of PUE samples with different flexibility of molecular chains was prepared by using 1,4-butanediol（1,4-BDO）/bisphenol-a（BPA） blends with different mole ratios including9/1, 8/2, 7/3, 6/4 and 5/5. As comparison, PUE extended with pure 1,4-BDO and BPA was also synthesized.These samples were characterized by differential scanning calorimetry（DSC）, thermogravimetric analysis（TGA）, dynamic mechanical analysis（DMA）, etc. The results showed that with the decrease in flexibility of molecular chains the glass transition temperature（Tg） increased and low-temperature properties became worse. Besides, all samples had a certain degree of microphase separation, and soft segments in some samples were crystallized, i.e. the decreasing flexibility of molecular chains led to the impossibility of chains tightly packing and crystalline domains forming so that the degree of microphase separation decreased and the thermal properties became worse.
文摘In this paper, a new theory for the dynamic properties of protein molecular system have been proposed by means of the analysis of the characteristics of the collective excitations generated under the localized fluctuation and the deformation of structure of the protein. Some new results obtained from this study show that the Davydov theory is an approximate and disadvantegeous theory.
基金Supported by the National Natural Science Foundation of China under Grant No 11004156the Natural Science Foundation of Shaanxi Province under Grant No 2014JM1025+2 种基金the Science and Technology Star Project of Shaanxi Province under Grant No2016KJXX-38the Special Foundation of Key Academic Subjects Development of Shaanxi Province under Grant No 2008-169the Xi'an Polytechnic University Young Scholar Supporting Plan under Grant No 2013-06
文摘Based on the nonequilibrium Green function method and density functional theory calculations, we theoretically investigate the effect of chirality on the electronic transport properties of thioxanthene-based molecular switch. The molecule comprises the switch which can exhibit different chiralities, that is, cis-form and trans-form by ultraviolet or visible irradiation. The results clearly reveal that the switching behaviors can be realized when the molecule converts between cis-form and trans-form. ~urthermore, the on-off ratio can be modulated by the chirality of the carbon nanotube electrodes. The maximum on-off ratio can reach 109 at 0.4 V for the armchair junction, suggesting potential applications of this type of junctions in future design of functional molecular devices.
基金Funded by the Talent Foundation and Youth Foundation of Xi’an University of Architecture and Technology(Nos.DB12062 and QN1239)
文摘Based on molecular mechanics and the deformation characteristics of the atomic lattice structure of graphene, a modifi ed molecular structure mechanics method was developed to improve the original one, that is, the semi-rigid connections were used to model the bond angle variations between the C-Cbonds in graphene. The simulated results show that the equivalent space frame model with semi-rigid connections for graphene proposed in this article is a simple, efficient, and accurate model to evaluate the equivalent elastic properties of graphene. Though the present computational model of the semi-rigid connected space frame is only applied to characterize the mechanical behaviors of the space lattices of graphene, it has more potential applications in the static and dynamic analyses of graphene and other nanomaterials.
基金Supported by the National Natural Science Foundation of China under Grant No 11374033
文摘We focus on two new 21） materials, i.e., monolayer and bilayer silicon phosphides （Sil P1）. Based on the elastic- scattering Green＇s function, the electronic-transport properties of two-dimensional monolayer and bilayer Au- Si1P1-Au molecular junctions are studied. It is found that their bandgaps are narrow （0.16eV for a monolayer molecular junction and 0.26 e V for a bilayer molecular junction）. Moreover, the calculated current-voltage char- acteristics indicate that the monolayer molecular junction provides constant output current （20 hA） over a wide voltage range, and the bilayer molecular junction provides higher current （42 hA）.
基金Supported by the National Natural Science Foundation of China under Grant Nos 11374033,11774030,51735001 and 61775016the Fundamental Research Funds for the Central Universities under Grant No 2017CX10007
文摘We investigate the electronic-transport properties of two-dimensional monolayer films from Au-P-Au molecular junction to Au-Si-Au molecular junction using elastic scattering Green＇s function theory. In the process of replacing the P atoms with Si atoms every other line from the middle of monolayer blue phosphorus molecular structure, the substitution of Si atoms changes the properties of Au-P-Au molecular junction significantly. Interestingly, the current value has a symmetric change as a parabolic curve with the peak appearing in Au-Si_1P_1-Au molecular junction, which provides the most stable current of 15.00 nA in a wide voltage range of 0.70-2.70 V.Moreover, the current-voltage characteristics of the structures indicate that the steps tend to disappear revealing the property similar to metal when the Si atoms dominate the molecular junction.
基金Supported by the National Natural Science Foundation of China(11102139,11472195)the Natural Science Foundation of Hubei Province of China(2014CFB713)
文摘Through molecular dynamics（MD） simulation, the dependencies of temperature, grain size and strain rate on the mechanical properties were studied. The simulation results demonstrated that the strain rate from 0.05 to 2 ns–1 affected the Young＇s modulus of nickel nanowires slightly, whereas the yield stress increased. The Young＇s modulus decreased approximately linearly; however, the yield stress firstly increased and subsequently dropped as the temperature increased. The Young＇s modulus and yield stress increased as the mean grain size increased from 2.66 to 6.72 nm. Moreover, certain efforts have been made in the microstructure evolution with mechanical properties association under uniaxial tension. Certain phenomena such as the formation of twin structures, which were found in nanowires with larger grain size at higher strain rate and lower temperature, as well as the movement of grain boundaries and dislocation, were detected and discussed in detail. The results demonstrated that the plastic deformation was mainly accommodated by the motion of grain boundaries for smaller grain size. However, for larger grain size, the formations of stacking faults and twins were the main mechanisms of plastic deformation in the polycrystalline nickel nanowire.
基金the Science Challenge Project(TZ2018004)the National Natural Science Foundation of China(21875228 and 21702195)for financial support。
文摘Finding energetic materials with tailored properties is always a significant challenge due to low research efficiency in trial and error.Herein,a methodology combining domain knowledge,a machine learning algorithm,and experiments is presented for accelerating the discovery of novel energetic materials.A high-throughput virtual screening(HTVS)system integrating on-demand molecular generation and machine learning models covering the prediction of molecular properties and crystal packing mode scoring is established.With the proposed HTVS system,candidate molecules with promising properties and a desirable crystal packing mode are rapidly targeted from the generated molecular space containing 25112 molecules.Furthermore,a study of the crystal structure and properties shows that the good comprehensive performances of the target molecule are in agreement with the predicted results,thus verifying the effectiveness of the proposed methodology.This work demonstrates a new research paradigm for discovering novel energetic materials and can be extended to other organic materials without manifest obstacles.
