139 results about "Molecular network" patented technology

The invention provides a large model lightweight reasoning deployment method under limited hardware resources, and the method comprises the steps: quantifying the weight importance of a large model through a composite index of gradient sensitivity and activation frequency, and carrying out pruning operation in combination with an improved index weighted moving average strategy, thereby obtaining a structured sparse model; the sparse model is divided into sub-networks by adopting double rules, a routing decision network is trained, and an adaptive feature shunting architecture model is constructed; a multi-precision weight set is generated through a nested quantization technology, quantization bit width is dynamically adjusted, and edge equipment hardware parameters are adapted to complete reasoning environment initialization; after a reasoning request is received, an optimal sub-network is selected based on the trained routing decision network, corresponding weights are loaded in parallel, and a reasoning result is fused and output; and converting a reasoning result format, and dynamically optimizing a scheduling strategy based on a system real-time monitoring index. The method is compatible with a mainstream large model and a hardware platform, and an efficient and universal deployment scheme is provided for end-side AI engineering landing.

The invention discloses a method for improving beef quality, an action mechanism and an experimental method, in a cattle body, vitamin A can activate the expression of EBF2 through an active metabolite RA of the vitamin A, and the EBF2 can inhibit the transcription process of CYP26B1 in a targeted manner to maintain the activity of a retinol signal channel, so that PPAR gamma and downstream lipid metabolism related genes thereof are activated, and the activity of the retinol signal channel is improved. Fatty acid transport and lipid accumulation in fat cells in the cattle muscle are promoted, fat deposition in the cattle muscle is promoted, and the beef quality is improved. By constructing the molecular network for regulating and controlling the formation of the fat in the cattle muscle, a complex regulation and control mechanism for controlling the fat deposition in the cattle muscle can be understood more deeply, a method for improving the beef quality based on regulating and controlling the vitamin A mediated key factor EBF2 is provided, and a new way is provided for improving the beef quality in a targeted manner.

The invention provides a multi-modal characterization molecular property prediction method based on hierarchical bidirectional cross attention, and relates to the technical field of machine learning assisted organic chemistry, and the method comprises the following steps: S10, generating same-molecule multiple sequences for data enhancement; s20, coding the sequence features through a pre-trained molecular language model MolBERT; s30, performing multi-modal feature fusion through a layered bidirectional cross attention mechanism; s40, establishing a prediction head; s50, in the reasoning stage, only the feature extraction and fusion steps are executed, and a molecular property prediction result is output through the trained prediction head. According to the method, the molecular sequence, the topological graph structure and the fingerprint features are effectively integrated, so that the prediction precision of the model on a plurality of MoleculeNet (molecular network benchmark) public data sets is superior to that of an existing method.

The invention provides a multi-channel adaptive communication controller based on an STM32 platform. The multi-channel adaptive communication controller comprises a user interaction module, a topology sensing module, a subnet division module, a channel sensing module, a route generation module, a data forwarding module and a working state monitoring module. The user interaction module is used for realizing interaction between a user and the communication controller; the topology sensing module is used for sensing the topology structure of a network where the communication controller is located in real time; if the communication controller automatically divides the subnets, the subnet dividing module divides the subnets according to the topological structure of the network acquired by the topology sensing module; the channel sensing module is used for monitoring the communication quality of a channel currently used by a network in real time; the data forwarding module is used for realizing a routing forwarding function; the working state monitoring module is used for monitoring the working state of the communication controller. Based on the topology sensing function, when the network topology changes, the routing can be adjusted in time, network delay and network congestion are avoided, meanwhile, a basis is provided for subnet division, and the workload of subnet configuration of a user is relieved.

The invention belongs to the technical field of honey analysis, and particularly relates to a method for analyzing characteristic chemical components of Qinling Mountain medium-peak honey. The analysis method comprises the following steps: obtaining different extraction layers, and carrying out mass spectrometry on each layer of sample; constructing a molecular network by mass spectrum data through a GNPS platform, and comparing the mass spectrum data with a database to obtain primary identification; unmatched nodes predict the skeleton by using an NAP tool to obtain a secondary result; then fractions are separated, and an unknown structure is deduced by combining an SMART technology with nuclear magnetism; and finally, synthesizing the three types of results to determine the characteristic chemical components of the honey. Compared with a single chromatography-mass spectrometry technology, the analysis method provided by the invention not only realizes data visualization, but also reduces the dependence of chemical component identification on a database to a certain extent, and provides reference for rapid characterization of chemical components in a honey crude extract or fraction.

The invention relates to the technical field of flexible stimulation response self-drivers, in particular to a preparation method of an electricity-temperature dual stimulation deformation reversible composite material driver, which comprises the following steps: firstly, carrying out chemical modification and solution ultrasonic dispersion on conductor particles to improve interface cohesiveness and improve interface bonding between a conductor and a polymer; melting, blending and compounding polyethylene vinyl acetate free radical double-group coupling by using a double-screw mixer to prepare a shape memory hemicrystalline and carbon-carbon single bond net-shaped cross-linked molecular network structure; uniformly doping and embedding conductor unit particles into a molecular network to construct a conductive three-dimensional network, and forming a conductive blend with a bicontinuous phase structure; and finally, thermally pressing the conductive blend to initiate a macromolecular free radical cross-linking reaction to prepare the electric-temperature dual-stimulation deformation reversible composite material driver with a conductive-semi-crystalline-cross-linked three-dimensional network structure. According to the invention, electric and thermal dual stimulation reversible deformation response of the shape memory composite material can be realized.

The invention relates to the technical field of functional liquid crystal materials, and discloses a liquid crystal film with an intelligent information encryption function and a preparation method and application thereof. The liquid crystal thin film comprises two layers of transparent substrates arranged in parallel and a composite functional layer between the transparent substrates. The composite functional layer comprises a honeycomb polymer matrix and liquid crystals filled in holes of the polymer matrix, and a polymer network is arranged in the liquid crystals; the distribution densities of the polymer networks in different areas of the liquid crystal film are different, so that the liquid crystal film has reversible information patterns which disappear (or display) during power-on and display (or disappear) after power-off; or the reversible change information pattern disappears (or is displayed) at a low temperature and is displayed (or disappears) at a high temperature. The distribution densities of polymer networks in different areas of the liquid crystal film are different, multiple responses of an electric field, a temperature, a magnetic field and the like are achieved, the visual effect that information patterns exist from nothing, exist from existence to nothing or change colors can be achieved, and therefore the intelligent information encryption function is achieved.

The invention provides a molecular force field modeling method and device, electronic equipment and a storage medium. The method comprises the following steps: establishing a fingerprint spectrum of each atom in a target molecule according to structural feature information and energy feature information of the target molecule; clustering the fingerprint spectrum of each atom in the target molecule to obtain a plurality of atom categories; modeling different types of atom clusters to obtain a potential energy equation between the atom clusters; and establishing a molecular force field according to potential energy equations of different types of atom clusters. The fingerprint spectrum of each atom in the target molecule is clustered, and the clustering process is effectively used for replacing the condition that the atom types are classified by human experience; a molecular network probability graph model is established through a Bayesian method, data mining is carried out, and multi-body energy is simulated, so that the dependence of a molecular force field on fitting data is reduced, and the accuracy of the simulated molecular force field is improved.

The invention relates to the technical field of thickening agents, and discloses a preparation process of a high-performance fracturing fluid emulsion type thickening agent. The preparation method comprises the following steps: S1, uniformly mixing 30-50% of oil phase and 50-70% of emulsifier; s2, preparing a water phase: dissolving a monomer and a cross-linking agent in water, and adjusting the pH to prepare the water phase; s3, emulsifying: dropwise adding the water phase into the oil phase, and performing high-speed shearing to form an inverse emulsion; and S4, polymerization reaction: adding an initiator under the protection of nitrogen, and carrying out polymerization reaction at 50-70 DEG C. By introducing the dynamic cross-linking agent, a reversible cross-linked network is formed in the molecular structure of the thickening agent. When the molecular chain of the thickening agent is broken under the action of high shearing force, the dynamic cross-linked structure can quickly reconstruct a molecular network, and the viscosity recovery rate of the thickening agent after shearing is ensured. And compared with the problem of low recovery rate caused by incapability of repairing broken molecules in the prior art, the adaptive capacity of the thickening agent in a high-shear environment is remarkably improved.

The present invention relates to the technical field of oilfield chemistry, and provides a wormlike supramolecular associated micelle, a hydrophobic associated copolymer and a preparation method thereof, the wormlike supramolecular associated micelle comprises a soft template, a hydrophobic functional monomer 1, a hydrophobic functional monomer 2 and a polarity regulator, the soft template is a long-chain molecular polymer and has an active functional group, and the polarity regulator is a hydrophobic functional monomer 2. The hydrophobic functional monomer 1 is an anionic-nonionic surfactant with polymerizable double bonds, the hydrophobic functional monomer 2 is a cationic-nonionic surfactant with polymerizable double bonds, and the polarity regulator is a micromolecular organic compound. The problems of poor resistance reduction, sand carrying capacity and the like of the fracturing fluid in the construction process due to dissociation of a polymer molecular network structure and sudden reduction of rheological property and viscoelasticity are solved.

The invention discloses a flexible packaging material and a preparation method and application thereof. The flexible packaging material comprises a sandwich structure formed by sequentially stacking a polymer elastomer with a viscoplastic surface layer, a self-adaptive gas barrier layer and a polymer elastomer with a viscoplastic surface layer, wherein the viscoplastic surface layer of the polymer elastomer consists of a non-polar viscoelastic molecular network and a plastic polar polymer microdomain; the self-adaptive gas barrier layer is at least one of liquid metal, hygroscopic gel and viscous hygroscopic liquid; and the viscoplastic surface layers of the upper and lower polymer elastomers are respectively contacted with two surfaces of the self-adaptive gas barrier layer, so that a sandwich structure is formed. The flexible packaging material not only can be flexibly stretched, but also has ultrahigh gas barrier property, and has important significance and value for promoting the practical application and industrialization process of flexible electronic devices.

The invention discloses a high-performance ionic piezoelectric hydrogel system based on molecular network reconstruction and micro-nano structure coordinated regulation and an establishment method thereof, and the establishment method comprises the following steps: S1, mixing acrylamide, polyacrylamide, sodium gluconate and polyvinyl alcohol to prepare a pre-polymerized solution, and adding HCl into the pre-polymerized solution to obtain a mixed solution system; and S2, carrying out gradient cross-linking and ultraviolet curing on the mixed solution system to obtain the high-performance ionic piezoelectric hydrogel system based on molecular network reconstruction and micro-nano structure coordinated regulation. The sensitivity is improved by 66%, and the problem that a traditional material is insufficient in response to weak biological signals (such as pulse micro-vibration) is solved.

The invention relates to a preparation method of molecular confinement fluorescent hydrogel and a fluorescence regulation method of the molecular confinement fluorescent hydrogel. According to the present invention, the intermolecular force drives the controllable self-assembly of the gel polymer chain to produce the compact advanced structure, and the self movement is effectively limited, such that the gel molecular network can produce the sequential fluorescence without the traditional conjugated chromophore so as to substantially simplify the fluorescent gel preparation process; meanwhile, multi-mode and large-range in-situ regulation and control are carried out on CIE system fluorescence based on high controllability of polymer self-assembly, so that the CIE system fluorescence has great functional application potential. The working principle of the CIE system is as follows: the aggregation state of the polyacrylic acid type polymer and the 1, 2-diamine type molecular unit is controllably converted under the driving of the characteristic intermolecular action, and the polymer chain can complete the conversion from the stretched form to the ordered aggregation form. The transition can greatly limit the movement of molecules in a highly aggregated state, thereby producing enhanced fluorescence emission.

The invention discloses yeast protein nutritional gel and a preparation method thereof, and belongs to the technical field of protein gel. Comprising the following steps: adding dialdehyde starch into a yeast protein solution, then sequentially adding EDC and NHS, fully mixing, and standing to obtain stable gel. The invention has the following technical effects: (1) the formation of a closer molecular network structure between YP and DS can be effectively promoted; (2) the mechanical property of the gel system is obviously improved; (3) an external structure is not introduced, and safety and environmental friendliness are achieved; and (4) rapid reaction can be realized under a mild low-temperature condition.

The invention discloses a method for screening lipid-lowering active substances from natural extracts by improving enzyme immobilization efficiency, compared with the prior art, a polypropylene hollow fiber membrane is used as an adsorption carrier, glutaraldehyde is used as a cross-linking agent, pancreatic lipase is covalently cross-linked to the polypropylene hollow fiber membrane, and the lipid-lowering active substances are screened from the natural extracts. The enzyme loading amount, the repeated utilization rate and the storage stability of the polypropylene hollow fiber membrane are obviously improved; the enzyme immobilization efficiency and the enzyme activity are improved; furthermore, liquid chromatography-quadrupole time-of-flight mass spectrometry and a molecular network analysis technology are combined to screen 31 active ligand structures from the wild chrysanthemum flower extract with numerous and jumbled components, so that the method can be popularized and used for efficiently screening active ligands in natural products with complex components; the material basis of the lipid-lowering effect of the natural-source active ingredients can be deeply illuminated, and the active ligand has certain lipid-lowering potential and can be used as a pilot chemical of lipid-lowering drugs.

The invention discloses an application of DUOX2 in identification of ileum subtype and colon subtype Crohn's disease. A group of new DUOX2-epi is analyzed and identified based on single cell transcriptome sequencing, the proportion of the cells in the Crohn disease inflammation tissue is obviously increased, and the proportion of the DUOX2-epi in the colon type Crohn disease inflammation tissue is obviously higher than that in the ileum type Crohn disease inflammation tissue; in addition, in ileum type and colon type Crohn diseases, interaction with surrounding cells is carried out through different mechanisms, so that the occurrence and development of the diseases are promoted. Therefore, based on a DUOX2-epi mediated intercellular communication network, a gene prediction module for predicting different subtypes of Crohn diseases and evaluating the treatment effect of a biological agent is developed, and based on a DUOX2 molecular network, an experimental basis and a research direction are provided for subsequent in-depth research, and a new strategy is expected to be provided for diagnosis and treatment of different subtypes of Crohn diseases.

The present application provides a molecular force field modeling method and apparatus, an electronic device, and a storage medium. The method comprises: establishing fingerprint spectrums of atoms in a target molecule on the basis of structural feature information and energy feature information of the target molecule; clustering the fingerprint spectrums of the atoms in the target molecule to obtain a plurality of atomic categories; performing modeling on different categories of atomic clusters to obtain a potential energy equation between the atomic clusters; and establishing a molecular force field according to the potential energy equation of different categories of atomic clusters. By clustering the fingerprint spectrums of the atoms in the target molecule, the clustering process is effectively used to replace the situation where atom types are classified on the basis of human experience; a molecular network probabilistic graphical model is established by means of a Bayesian method, data mining is carried out, and multi-body energy is simulated, so that the dependence of a molecular force field on fitting data is reduced, and the accuracy of a simulated molecular force field is improved.

The invention belongs to the technical field of biological information, and particularly relates to a cancer molecular subtype recognition method based on self-adaptive pellet multi-view image clustering. The method comprises the following steps: acquiring a multi-omics data set of cancer molecules, and constructing a pellet set for each kind of omics data in the multi-omics data set; constructing a biological network structure chart according to the particle ball set; fusing the biological network structure diagrams of the omics data to obtain a unified graph; inputting the unified graph into a pre-trained heterogeneous graph neural network for processing to obtain a cancer molecular subtype recognition result; according to the method, multi-scale biological structure features in multiple omics data can be captured at the same time, collaborative optimization of molecular network topology and patient characterization is achieved, and therefore the accuracy of cancer molecular subtype recognition results is improved.

The invention discloses a low-odor regenerated polypropylene composite material as well as a preparation method and application thereof, and belongs to the technical field of polymer composite materials. Comprising the following raw materials in parts by mass: 10-78 parts of polypropylene, 10-50 parts of regenerated polypropylene, 0-20 parts of an inorganic filler, 2-15 parts of a toughening agent, 0.5-2 parts of an isocyanate coupling agent, 1-3 parts of an odor adsorbent, 0.1-0.5 part of an initiator and 0.5-2 parts of a processing aid. By adding an isocyanate coupling agent, in the extrusion granulation process, an isocyanate group (-NCO) and hydroxyl (-OH) or amino (-NH2) or carboxyl (-COOH)-containing micromolecules remaining in regenerated polypropylene or generated in the extrusion process are subjected to an addition reaction to form urea bonds or urethane bonds, and a plurality of micromolecules are cross-linked to form a macromolecular network; the small molecules with odor are captured, so that the content of emitted substances is reduced; in addition, small molecules are further adsorbed by cooperatively utilizing an odor adsorbent, and the low-emission regenerated polypropylene composite material is prepared.

The invention discloses a corrosion-resistant ABS plastic and a preparation method thereof, and relates to the technical field of ABS preparation, the corrosion-resistant ABS plastic comprises the following raw materials by weight: 80-100 parts of modified ABS, 5-8 parts of modified graphene and 0.02-0.05 part of a Karst catalyst, in the melt extrusion process of the modified ABS, the modified graphene and the Karst catalyst, the modified ABS, the modified graphene and the Karst catalyst are subjected to melt extrusion to obtain the corrosion-resistant ABS plastic. Si-H bonds on the modified ABS and double bonds on the modified graphene are grafted to further form more crosslinking points, so that a more compact three-dimensional network structure is formed between a modified ABS molecular chain and the modified graphene, a two-dimensional lamellar structure of the modified graphene can form a uniform covering layer in a modified ABS matrix, and a fluorine-containing polysiloxane chain segment is embedded in a molecular network, so that the fluorine-containing polysiloxane-modified ABS composite material has the advantages that the fluorine-containing polysiloxane-modified ABS composite material is more excellent in heat resistance and heat resistance, and the service life of the fluorine-containing polysiloxane-modified ABS composite material is prolonged. The surface tension of the material surface is obviously reduced, the contact between a corrosive medium and the material surface is reduced, and the corrosion resistance of the material is further improved.

The invention belongs to the field of rape breeding, and particularly discloses a sclerotinia rot resistance related gene BnaCLE20 of rape and application of the sclerotinia rot resistance related gene BnaCLE20 in rape breeding. The sequence of the BnaCLE20 gene is as shown in SEQ ID NO. 1, and the coded protein is as shown in SEQ ID NO. 2; the BnaCLE20 gene of the brassica napus is identified to be capable of regulating the sclerotinia rot resistance of the brassica napus for the first time, a molecular network for regulating the sclerotinia rot resistance of plants is enriched, and the BnaCLE20 gene plays an important role in researching a molecular mechanism formed by the disease resistance of the plants. The overexpression of the BnaCLE20 gene disclosed by the invention can reduce the content of active oxygen, malondialdehyde and oxalic acid in rape leaves infected by sclerotinia sclerotiorum; meanwhile, the activity of catalase in the rape leaves infected by sclerotinia sclerotiorum is enhanced; the sclerotiniose resistance of the oilseed rape is remarkably improved, and the method has important significance in guaranteeing the high and stable yield of the oilseed rape by improving the sclerotiniose resistance of the oilseed rape.

The invention discloses a flexible packaging film, a preparation method thereof and a flexible electronic device. Wherein the flexible packaging film takes an elastomeric polymer with a flexible chain segment and a rigid chain segment as a base material, the flexible packaging film comprises an elastomeric polymer surface and a viscoplastic surface, and hard-phase polar polymer microdomains are discretely distributed in the viscoplastic surface. The viscoplastic surface comprises a non-polar elastomer polymer viscous molecular network and a plastic polar polymer micro-domain, the polar polymer micro-domain and the viscous molecular network have good compatibility, interface interaction is enhanced to form a defect-free interface, the film is endowed with excellent adhesion capacity, and the water vapor barrier efficiency is improved. Meanwhile, the hard-phase polar polymer microdomains are only distributed on the viscoplastic surface on one side of the film, so that the mechanical property of the film is ensured, and the film can be tightly bonded with various matrixes and also has high stretchability and high barrier property.

The invention discloses a yellow rice wine quality detection and analysis method, which comprises the following steps: (a) grouping a plurality of yellow rice wine samples according to whether caramel color is added or not, and the years in each group are the same in a one-to-one correspondence manner; a plurality of parallel samples are taken from each wine sample, detection is performed through a QTOF high-resolution mass spectrometer, effective information of TOF MS and TOF MS / MS is obtained through one-time sample introduction in a DBS-IDA collection mode, and high-resolution first-level and second-level mass spectrum data are obtained; (b) analyzing by adopting SCIEX OS software, and identifying a plurality of compounds according to an SCIEX database and literatures; the structure is identified through a GNPS molecular network technology, and SCIEX OS is used for further verification and confirmation of a plurality of other compounds; and (c) carrying out difference analysis through PLS-DA. According to the method, the influence of the caramel color additive and the aging year on the yellow wine quality can be scientifically and comprehensively analyzed, and a basis is provided for yellow wine quality evaluation and quality control.

The invention relates to picrasma quassioides bitter taste element compounds in Chinese picrasma quassioides as well as a preparation method and application thereof, belongs to the technical field of medicines, and particularly relates to 19 picrasma quassioides bitter taste element compounds obtained from branches and leaves of the Chinese picrasma quassioides on the basis of a molecular network annotated with Mzmine and DeepMass from picrasma quassioides which belongs to picrasma quassioides. They structures and configurations are determined by spectral analysis, quantum chemistry calculations, and comparisons of experimental ECD and computational data. In addition, the food refusal activity of all the compounds to plutella xylostella has been assessed, and the food refusal activity of all the compounds to plutella xylostella has been assessed. Wherein part of the compounds show medium antifeedant activity to plutella xylostella. The preparation method is simple and good in reproducibility, and the obtained compound is high in purity.

This invention provides a system and program that facilitates the process of generating and / or visualizing molecular networks for library molecular structure content. [Solution] In system 500, the molecular network generation system 502 for library molecular structure content includes an evaluation component 512 that performs a comparison between a first molecular fingerprint including first molecular structure data of a first molecular structure and a second molecular fingerprint including second molecular structure data of a second molecular structure, and a visualization component 516 that generates display data that visualizes a representation of the structural similarity score obtained from the first molecular structure, the second molecular structure and the comparison. The representation by the visualization component includes edges corresponding to the structural similarity score that extend between pairs of nodes corresponding to the first molecular structure and the second molecular structure.

The invention discloses a molecular network modified composite solid electrolyte as well as a preparation method and application thereof. The composite solid electrolyte is prepared from the following raw materials in percentage by weight: 5wt%-30wt% of solute salt, 5wt%-80wt% of inorganic filler and 10wt%-80wt% of a polymer matrix, and the polymer matrix consists of a compound A, a polymer B and a polymer C in a mass ratio of 10: 1: (1.5-4.5). The invention also specifically discloses a preparation method of the composite solid electrolyte and an application of the composite solid electrolyte in preparation of a lithium metal battery. The composite solid electrolyte prepared by the invention has high ionic conductivity and smooth morphology, and can be fully contacted with an electrode to realize stable circulation when being applied to a lithium metal battery.

This application relates to the field of polymer materials technology, providing a PET elastomer, composite material, dynamic covalent network polymer, and its preparation method and application. The elastomer is prepared by in-situ self-assembly of crown ether-primary amide dual-functionalized PET through melt blending. Its molecular network integrates a quasi-rotaxane supramolecular crosslinking network. This structure is formed by the host-guest interaction between the crown ether and the primary amide, and can act as a sliding crosslinking point to dissipate energy through the sliding of cyclic components. This application employs a two-step end-capping-in-situ self-assembly process to overcome the technical obstacles of low PET chain-end activity and poor compatibility between polyrotaxane and PET, achieving molecular-level dispersion of quasi-rotaxane in the PET matrix. This fully utilizes the molecular pulley effect to achieve breakthrough toughening of the elastomer. Simultaneously, it combines rigid nanofillers modified with aminosilane coupling agents to form a soft-hard synergistic network, enabling the PET composite material to maintain high toughness while simultaneously improving strength, modulus, and thermal properties.

The invention relates to the technical field of medical biology, and discloses a traditional Chinese medicine quality marker screening method and an angelica dahurica variety quality marker group, the traditional Chinese medicine quality marker screening method and the angelica dahurica variety quality marker group comprise the following steps: respectively crushing, extracting and detecting a newly bred angelica dahurica variety and a farm variety; carrying out pretreatment on mass spectrum data by adopting MZmine 2.5 software; and the coumarin components are screened by adopting a multi-stage mass loss window. According to the traditional Chinese medicine quality marker screening method and the angelica dahurica variety quality marker group, through integration of ultra-high performance liquid chromatography-tandem mass spectrometry and polygon quality defect filtering, extended virtual molecular formula library screening, characteristic molecular network, network annotation transmission and other calculation strategies, a plurality of coumarin components are successfully annotated; and multivariate statistical analysis is adopted, so that the quality marker of the newly cultivated angelica dahurica variety is obtained.

The invention discloses a drug collaborative screening system based on artificial intelligence prediction and metabonomics verification, and belongs to the technical field of artificial intelligence and bioinformatics crossing. Comprising a multi-source data acquisition and feature fusion module, an artificial intelligence prediction module, a drug screening and visualization module, a metabonomics verification module and a feedback optimization and self-learning module. According to the invention, through multi-modal data integration and feature coding, unified modeling of a drug-disease-molecule network is realized; drug combination synergistic effect prediction is realized through an artificial intelligence algorithm; a prediction result is verified by using non-targeted metabonomics data, and the mechanism interpretability of the model is enhanced; a technical system capable of realizing closed-loop optimization is constructed, and an algorithm and experiment dual-verification platform is provided for drug discovery and precision medicine.

The invention discloses a method for automatically identifying new pollutants and structural analogues thereof in environment complex samples, which comprises the following steps of: firstly, analyzing co-occurrence of abundance of substances among samples by using a Pearson's correlation coefficient, and introducing program blank deduction and IQR-Fisher Z transformation abnormal value processing to improve reliability; then, a molecular network is constructed based on high correlation and spectrogram similarity, and cracking products in the source are filtered. Key points are as follows: an iterative annotation strategy is adopted: a candidate structure is obtained from a PubChem database according to m / z of a substance to be annotated, preliminary screening is performed by calculating a Tanimoto coefficient of a Morga fingerprint, or advanced annotation is performed in combination with predicted MS / MS spectrogram fragment matching of a Top-k candidate, and the substance which is successfully annotated is taken as a new round of seed compound, so that automatic recursive expansion of an annotation range is realized. According to the method, the dependence of a traditional method on a standard spectrogram is effectively overcome, and efficient and automatic identification of pollutants with novel structures and analogues thereof is realized.