11 results about "Atomic coordinates" patented technology

The application relates to a multi-descriptor and coordinate reversible mapping atomic clustering analysis method and system, which comprises the following steps: inputting a set of multiple descriptors representing atomic arrangement characteristics; selecting single descriptor data or comprehensive descriptor data obtained by multi-descriptor weighted combination according to system characteristics by a user; performing clustering analysis on the selected descriptor data, and generating descriptor data grouping and a sample index set corresponding to each group; extracting the Coulomb matrix of each sample in each group from the set of multiple descriptors based on the sample index set of each group, forming a Coulomb matrix subset of each group and respectively calculating the Coulomb matrix mean value; and applying a multi-dimensional scaling algorithm to reduce the Coulomb matrix mean value of each group to a three-dimensional space to generate corresponding three-dimensional atomic coordinates of each group. Compared with the prior art, the application improves the analysis capability of atomic or molecular arrangement characteristics in the data of simulation outputs such as molecular dynamics.

The application relates to a training method and device of a crystal property prediction model and a computer device. The method comprises the following steps: obtaining a sample crystal and a property label corresponding to the sample crystal; the sample crystal comprises a plurality of sample atoms; determining respective atomic coordinates of each sample atom, and determining orbital characteristics of the sample atoms according to the atomic coordinates; determining respective atomic potential energies of the sample atoms according to the orbital characteristics, and determining a predicted property corresponding to the sample crystal according to the atomic potential energies; training the crystal property prediction model according to the difference between the predicted property and the property label, stopping when a first training stop condition is reached, and obtaining a trained crystal property prediction model; and the trained crystal property prediction model is used to determine the property of the crystal. The method can improve the accuracy of predicting the property of the crystal.

The application provides an electron density information acquisition method, system, medium, terminal and program product based on machine learning. The method comprises the following steps: first, constructing a real differential density according to a real electron density and a reference electron density; inputting atomic species, atomic coordinates, cell information and uniform grid definition information of a target structure into a machine learning model for training, and taking the real differential density as a supervision target during the training to construct a differential density prediction model; deploying the constructed differential density prediction model to output a predicted differential density of a to-be-tested complex structure according to to-be-tested complex structure information; and performing a recovery operation on the output predicted differential density to obtain total electron density information of the to-be-tested complex structure. The application can acquire electron density information of a complex structure with high precision and high efficiency, and stably support downstream calculations such as photoelectric properties and quantum transport of heterostructures, interface structures and device structures.

This invention discloses a molecular dynamics calculation method and system based on quantum simulation. The method first applies a discrete quantum Fourier transform to the atomic coordinates of the molecular system, mapping the original plane wave basis to a system containing only O(N) atoms. 2 The invention constructs a plane-wave bibasic quantum mechanics equation; then, it constructs and solves the electronic structure Hamiltonian on a quantum computer to obtain the ground state energy. This quantum energy is combined with the classical empirical potential energy according to predetermined weighting coefficients to calculate the total synthesized energy, from which the atomic forces are determined. The integration time step is then adaptively adjusted based on a comparison of the quantum energy change between consecutive time steps with a preset threshold. This adaptive mechanism shortens the time step to ensure accuracy when quantum effects are significant, and lengthens the step to improve efficiency when changes are gradual. This invention significantly reduces Hamiltonian complexity, balancing quantum accuracy and classical efficiency, and is applicable to various fields such as drug screening, materials design, and catalytic reaction simulation.

The application discloses a three-dimensional molecular structure generation method and model, relates to the technical field of the cross of computational chemistry and machine learning, and the three-dimensional molecular structure generation method comprises the steps of: generating an initial three-dimensional conformation by adopting a preset distance geometry algorithm in combination with a preset molecular force field optimization based on molecular topological information; constructing three SE(3) isovariant geometric graphs according to the initial three-dimensional conformation and the molecular topological information, and generating isovariant node representations of the graphs based on the SE(3) isovariant geometric graphs; fusing the isovariant node representations of the SE(3) isovariant geometric graphs to obtain a molecular embedding; predicting atomic coordinate offsets according to the molecular embedding and preset DFT true value coordinates; and generating a target three-dimensional molecular structure according to the initial three-dimensional conformation and the atomic coordinate offsets. The application can realize high-precision, high-efficiency and strong-robustness three-dimensional conformation prediction under limited computing resources and data scale.

This invention discloses a heterogeneous node electrostatic force calculation device for molecular dynamics simulations, comprising: a CPU node for constructing molecular topology, calculating Gaussian width parameters, initializing short-range force lookup tables, and outputting atomic trajectories and system energy after simulation; an FPGA node connected to the CPU node via a PCIe bus for selecting short-range atom pairs based on truncation radius and atomic coordinates, and calculating short-range electrostatic forces in parallel using a lookup table; and a GPU node connected to both the FPGA and CPU nodes via the same PCIe bus for performing long-range electrostatic force calculations on all atoms in the system, and updating atomic positions and velocities after reducing short-range and long-range forces. The heterogeneous node electrostatic force calculation device for molecular dynamics simulations provided by this invention, through reasonable task allocation, fully utilizes the advantages of GPUs and FPGAs to achieve asynchronous parallel computing between GPUs and FPGAs, thereby improving the overall speed of electrostatic force calculations.

The application discloses a protein essentiality prediction method based on multi-modal double fusion, comprising the following steps: 1, obtaining data containing essential protein sequences and non-essential protein sequences and preprocessing; 2, using a large language model to extract residue embedding matrix, residue contact matrix and three-dimensional atomic coordinates from the preprocessed data; 3, extracting each modal information from the large language model through a feature processing module; 4, realizing the fusion of one-dimensional sequence and two-dimensional structure features based on a fuzzy self-adaptive clustering fusion module; 5, realizing the fusion of one-dimensional sequence and three-dimensional structure features based on a bidirectional co-attention fusion module; 6, constructing a protein essentiality classification layer network to realize the prediction of protein essentiality. The application solves the problems of single feature representation and insufficient sequence and structure feature fusion in the traditional protein essentiality prediction method, and improves the accuracy and robustness of protein essentiality prediction.

This invention relates to the field of computer-aided process design technology, specifically to a method and system for optimizing the formulation of biodegradable tableware materials. The method includes the following steps: generating equilibrium molecular conformations based on atomic coordinates and calculating initial formulation component parameters; converting modulus parameters to generate theoretical elastic recovery strain parameters; mapping models to calculate deformation and marking diffusion characteristic length parameters; calculating diffusion coefficients and flux parameters to update local hydrogen ion concentration indices; updating reaction rate constants and integrating them to generate finalized formulation structural parameters. In this invention, a microscopic molecular dynamics model is constructed to analyze interface voids to quantitatively assess compatibility; relaxation time spectra are used to invert elastic recovery strain to predict deformation risk; autocatalytic reaction and diffusion theories are combined to track hydrogen ion concentration evolution; a dynamic mapping between degradation and failure is established; and full life-cycle verification is completed in a virtual environment, significantly shortening the R&D cycle and reducing trial-and-error costs.

This invention provides a molecular identification and extraction method based on ReaxFF reaction kinetics simulation, belonging to the field of molecular identification and extraction technology. To obtain molecular structure information from reaction kinetics simulations, this invention obtains a dump source file containing bond orders and coordinates from the ReaxFF simulation output. It stores the bond order information at each time step in the bond source file, extracts and stores the atomic information of all molecules at the same time step, determines the target molecule features, identifies and stores the target molecule features based on the obtained atomic ID information of all molecules at the same time step, stores the coordinate information of the dump source file at each time step, and iterates through the atomic coordinate information and atomic IDs of all target molecules at the same time step to obtain the atomic coordinate information of the target molecule. This invention is applied to the identification and extraction of molecular structure information.

The application relates to the technical field of material modeling, and discloses a modeling method for the synergistic action of multiple hydration products based on molecular dynamics simulation, which comprises the following steps: based on the atomic coordinates of the hydration products, determining a regulation range and a proportion range, and constructing an initial solid mixture model by using charge balance and water molecule adding mode; configuring simulation environment parameters and force field values for the initial solid mixture model, and executing mechanical structure optimization and dynamics simulation of the initial solid mixture model based on the configuration results, and outputting a stable solid mixture model based on the simulation results; extracting target parameter information of the stable solid mixture model, comparing the target parameter information with experimental standard data, verifying the deviation state of the stable solid mixture model according to the comparison results, optimizing the stable solid mixture model based on the deviation state, and outputting a stable solid mixture model for describing the synergistic action of the hydration products until the deviation state meets the requirements. The application provides direct theoretical support for mechanical property optimization through the analysis of the bonding state and atomic diffusion behavior at the nanoscale level.

This application relates to a molecular optimization method and system based on protein structure and affinity guidance. The method includes: inputting structural data of protein pockets and the molecular backbone into a diffusion model; in the forward process, noise is progressively introduced from a Gaussian distribution at each time step, and prior guidance features of the R-group are extracted using a pre-trained prior guidance model, with the mean shift of the atomic coordinates of the R-group calculated and added to the forward process; in the reverse process, sampling from the Gaussian distribution is used as a starting point, and the R-group matching the molecular backbone is generated through progressive denoising at each time step, based on the protein pockets and the molecular backbone; simultaneously, at each time step, a guidance correction term is calculated based on the prior guidance features to guide the correction of the atomic coordinates of the R-group generated at the current time step; after denoising, the molecular backbone and the R-group are assembled into a complete molecule and output. This invention can improve the binding affinity between molecules and protein pockets.