44 results about "Quantitative structure" patented technology

The invention discloses a bidirectional reversible conversion method and system between a peptide molecule SMILES and a sequence expression. The core innovation lies in that a new sequence description syntax is defined to retain information of a polypeptide special bond and specific modification of amino acid; a main chain atom index and adjacency traversal topology identification algorithm is adopted, and end group and topology integrated detection and coding are carried out; a residue recognition algorithm for main chain cutting and template library matching is compatible with any standard or non-standard amino acid residues, an extensible end group library / monomer template library and an automatic increment mechanism, and automatic recognition and sequence annotation of S-S disulfide bonds; the invention relates to a high-fidelity assembly algorithm of HELM anchor points and topology aware cyclic peptide processing. The method solves the problems of incapability of supporting a complex polypeptide topological structure, poor reversibility, insufficient expansibility of a monomer library and the like in the prior art, can be widely applied to scenes of quantitative structure-activity relationship model construction, large-scale polypeptide data cleaning and the like, and has remarkable practicability and innovativeness.

The invention discloses a rare and endangered organism toxicity prediction method based on a machine learning algorithm and a quantitative structure-activity relationship, which constructs a toxicity prediction model through the machine learning algorithm, can effectively assess the toxicity influence of environmental pollutants on rare and endangered organisms, and provides technical support for rare and endangered organism protection and ecological environment risk assessment. Comprising the following steps: 1, collecting data related to rare and endangered organisms from a public database, and establishing a rare and endangered organism toxicity prediction database based on machine learning; 2, generating molecular descriptors for the chemical substances; step 3, data preprocessing; 4, development of acute and chronic toxicity prediction models of rare and endangered organisms based on multiple machine learning is carried out, and performance evaluation is carried out; 5, performing internal and external verification on the rare and endangered biotoxicity prediction model; step 6, analyzing the importance of the features by using the valuable, rare and endangered biotoxicity prediction model, and finding out the most influential features; and 7, predicting the toxicity value of the pollutants in combination with the optimal machine learning model.

The present disclosure relates to systems and methods for chemical toxicity prediction. The methods of the present disclosure comprise: receiving test data from an analysis instrument; selecting a candidate chemical according to the test data; determining a hazard translated level and a hazard evaluation level of the candidate chemical according to a molecular fingerprint of the candidate chemical; and predicting the toxicity of the candidate chemical by using a quantitative structure-activity relationship (QSAR) model based on the hazard translated level and the hazard evaluation level.

The invention discloses an ecological risk assessment method for organic pollutants in a natural water body, and belongs to the field of ecological risk assessment of the natural water body, and the method comprises the steps: building a quantitative structure-activity relationship toxicity prediction model of species based on a machine learning and feature screening method; predicting and obtaining biotoxicity data in a to-be-predicted water body by using the quantitative structure-activity relationship model, establishing a species sensitivity distribution model of pollutants in the water body according to the biotoxicity data, and calculating predicted non-effect concentration; on the basis of database retrieval, acute toxicity experiments and an inter-species relationship prediction model, collecting, predicting and supplementing toxicity data of species in the water body, and evaluating the harmfulness of pollutants in the water body by using a risk entropy method and a comprehensive scoring method respectively. According to the method, ecological risk assessment is carried out by utilizing a machine learning method, the toxicity of the persistent organic pollutants in seawater and fresh water can be effectively assessed, and the ecological risk of the organic pollutants in the water body can be effectively assessed.

The invention belongs to the technical field of biological information, and relates to an acetylcholin esterase inhibitor prediction method based on Stacking ensemble learning and molecular feature fusion, which comprises the steps of data collection and preparation, data annotation and optimization, feature extraction and analysis, construction of a Stacking model, result verification and feedback and construction of a prediction platform. The molecular fingerprints and the property descriptors are used as features, and an acetylcholin esterase inhibitor classifier is successfully constructed by adopting a Stacking algorithm. According to the method, the problems that the efficiency of finding the acetylcholin esterase inhibitor by a traditional experimental method is low, and a common quantitative structure-function relationship method is high in complexity and poor in generalization ability can be solved, the new drug finding speed is increased, experimental candidates are accurately positioned, and resource waste is reduced.

The invention provides a glass elasticity modulus prediction method and system based on molecular dynamics simulation, and belongs to the technical field of glass performance prediction.The glass elasticity modulus prediction method comprises the steps that glass materials of different systems are obtained and tested, and an elasticity modulus test database is constructed; constructing an atomic model containing different system glass material component atoms, and optimizing the model; performing molecular dynamics simulation calculation on the optimized atomic model to obtain structure information of glass materials of different systems; enabling the elasticity modulus test databases of the glass materials of different systems to correspond to the structural information, and constructing a component-structure-performance data set; the elastic modulus is used as a dependent variable, the structural information is used as an independent variable, a multiple regression model is constructed, a quantitative structure-activity relationship for predicting the modulus of the glass elastic model is established, and the elastic modulus of a glass material in actual experimental production is predicted through simulation of molecular dynamics on the structural information. According to the invention, accurate simulation of the glass system structure and accurate prediction of the elastic modulus are realized.

The invention provides a method for predicting the flammability upper limit volume percentage of a pure compound. The method can be used for predicting a mathematical model of the flammability upper limit volume percentage of the pure compound which is composed of 12 or less elements of hydrogen, carbon, nitrogen, oxygen, sulfur, fluorine, chlorine, bromine, iodine, silicon, phosphorus, arsenic and the like and has the atom number of 25 or less (excluding hydrogen) with high accuracy. Wherein the model is obtained as a universal quantitative structure-property relationship model by determining an optimal model from a plurality of multiple linear regression models by means of a step-by-step selection method, and the model takes some of the various molecular descriptors as independent variables, takes flammability upper limit volume percent as a dependent variable, and takes the flammability upper limit volume percent as a dependent variable. The value of the molecular descriptor included in the model can be received and input in a short time, and the flammability upper limit volume percentage can be output, so that only the specific value of the molecular descriptor included in the model is known. And the flammability upper limit volume percentage of the compound purely formed by the molecule can be predicted for any molecule meeting the requirements of the invention.

Provided is a mathematical model that can predict, with high accuracy, the saturated liquid density of a pure compound at 298.15 K, said pure compound comprising 12 or less elements such as hydrogen, carbon, nitrogen, oxygen, sulfur, fluorine, chlorine, bromine, iodine, silicon, phosphorus, and arsenic and having 25 or less atoms (excluding hydrogen). The model is a quantitative structure-property relation model and is obtained by solving an optimal model from a plurality of multiple linear regression models by a step-by-step selection method, and the model takes some molecule descriptors in various molecule descriptors as independent variables, takes saturated liquid density under 298.15 K as a dependent variable, and takes the saturated liquid density under 298.15 K as a dependent variable. The value of the molecule descriptor included in the input model can be received in a short time, and the saturated liquid density at 298.15 K can be output, so that the saturated liquid density at 298.15 K of a compound formed by the molecule singly can be predicted for any molecule meeting the requirements of the invention as long as the specific value of the molecule descriptor included in the model is known.

The invention discloses a method for predicting combined toxicity of a graphene-based material and ionic liquid. The method comprises the following steps: obtaining an adsorption coefficient through an adsorption experiment of the graphene-based material on the ionic liquid; establishing a concentration-reaction relationship of a binary mixture composed of the graphene-based material and the ionic liquid to a specific ecological species, and obtaining a mean effect concentration of the binary mixture; fitting the adsorption coefficient and the mean effect concentration through a machine learning algorithm to obtain a fitting relationship, and obtaining a quantitative structure-activity relationship model according to the fitting relationship; and inputting a to-be-detected adsorption coefficient serving as an input variable into the quantitative structure-activity relation model to obtain the combined toxicity of the binary mixture on the specific ecological species. The adsorption coefficient is used as the characteristic parameter corresponding to the toxicity data, so that the characteristic parameter is easier to obtain, and the time, manpower, material and other costs caused by biological testing when the average effect concentration is obtained are avoided; and through a machine learning method, a fitting relation is constructed, and prediction can be more efficient and accurate.

The invention provides an ecological risk evaluation method and system based on new pollutants, and the method comprises the steps: obtaining mass spectrum data of sample data, determining a normal mass spectrum fingerprint corresponding to a geographic position and time, marking an ion peak as an unknown peak when the ion peak in the mass spectrum data deviates from a preset range of the normal mass spectrum fingerprint, and determining that the ion peak is an unknown peak; the method comprises the following steps: acquiring a fragmented spectrum of an unknown peak based on secondary mass spectrometry, inferring a possible molecular structure of the unknown peak through a graph neural network model in combination with a preset chemical database, and constructing a quantitative structure-activity relationship of possible molecular structure analogues of the unknown peak based on a preset structural similarity network. The method comprises the following steps: acquiring predicted toxicity and signal channels influenced by the predicted toxicity, performing ecological risk assessment in a preset range of a petrochemical plant based on the predicted toxicity and the signal channels influenced by the predicted toxicity, and acquiring potential ecological risks of new pollutants. And a basic basis is provided for environment management and decision making.

Systems, methods, and / or apparatuses may be operative to perform a dialysis process that includes a displacer infusion process. In one embodiment, a method for determining a displacer compound may include constructing a plurality of target protein quantitative structure-activity relationship (QSAR) models, one for each of the plurality of binding sites, analyzing a set of candidate compounds using the plurality of QSAR models to determine a set of at least one potential compound with an affinity for binding to each of the plurality of binding sites, and selecting at least one displacer compound from the set of at least one potential compound. Other embodiments are described.

1. The name of the design product: quantitative structure-activity relationship prediction graphical user interface of electronic equipment. 2. The use of the design product: an electronic device. 3. The design points of the design product: the interface content of the graphical user interface. 4. The picture or photo that best indicates the design points: front view. 5. The use of the graphical user interface: to show the details of the prediction task of the correlation between the quantitative structure-activity relationship chemical structure and biological activity, the front view is the user task management interface, clicking the icon button under the prompt of Actions in the prediction task row, entering interface change state figure 1, clicking the View model parameters button on the interface, entering interface change state figure 2.

1. Name of this design product: Quantitative structure-activity analysis prediction graphical user interface for electronic devices. 2. Purpose of this design product: an electronic device. 3. The key design point of this design product lies in the interface content of the graphical user interface. 4. The picture or photo that best illustrates the key points of the design: main view. 5. For conventional design, the rear view, left view, right view, top view and bottom view are omitted. 6. Purpose of the Graphical User Interface: Used to display the details of the prediction task for the correlation between chemical structure and biological activity in quantitative structure-activity analysis. The main view is the user task management interface. Click the icon button with the prompt "Details" under the Actions column of the prediction task row to enter the interface change state diagram 1, which displays the prediction details. Scroll to the bottom of the interface to enter the interface change state diagram 2, which further displays the prediction details.

The invention provides a method for predicting a flash point of a pure compound by using a multiple linear regression model. According to the method, a mathematical model for predicting the flash point of the pure compound with high accuracy is established to predict a flash point value. Specifically, the model is used as a quantitative structure-property relationship model, after the optimal model is obtained from a plurality of multiple linear regression models by using a stepwise selection method for most compounds with known flash point experimental values, the molecular descriptors included in the model are received in a short time, and the flash point values are output. As long as the specific values of the molecule descriptors included in the model are known, the flash point of the compound composed of the molecule is predicted for any molecule. As described above, the present invention provides a method and model capable of predicting a reliable flash point value even for many compounds having unknown experimental values, thereby saving the cost and time for physical performance measurement experiments or chemical structure prediction, and enabling research and development activities of related industries to become easier.

The invention discloses a method for directionally screening a factor G for QSTR (Quantitative Short Tandem Resonance) analysis. The method comprises the following steps: 1) selecting an isotropic compound to be screened; 2) respectively carrying out friction performance test on the to-be-screened compounds to obtain corresponding friction performance data; (3) calculating a G value of a compound to be screened by utilizing a calculation formula of a screening factor G and combining the obtained friction performance data; drawing a G value distribution diagram based on different to-be-screened compounds; and 4) determining a critical value in the G value distribution diagram according to the critical value percentage S%, and screening compounds with G values greater than the critical value as target molecules. According to the method, key samples suitable for entering quantitative structure-tribological performance relationship QSTR analysis can be quickly and accurately screened out from a large number of material candidate libraries, the purpose of quickly and accurately screening compounds with target effects is achieved, the screening efficiency and accuracy are remarkably improved, and organic fusion of QSTR and tribological test analysis is promoted.

1. The name of the design product: quantitative structure-activity relationship prediction graphical user interface of electronic equipment. 2. The use of the design product: an electronic device. 3. The design points of the design product: the interface content of the graphical user interface. 4. The picture or photo that best indicates the design points: front view. 5. The use of the graphical user interface: to show the details of the prediction task of the correlation between the quantitative structure-activity relationship chemical structure and biological activity, the front view is the user task management interface, clicking the icon button under the prompt of Details in the Actions column of the prediction task row enters the interface change state figure 1, and in the interface, clicking the picture in the SMILES card enters the interface change state figure 2.

The invention belongs to the technical field of measurement, and relates to a qualitative and quantitative analysis method for drinking water disinfection by-products based on effect orientation. According to the method, a drinking water sample is separated in a grading manner, high-toxicity components are screened out in combination with cytotoxicity evaluation, disinfection by-products in the high-toxicity components are identified in a non-targeted manner by utilizing a high-resolution liquid chromatography-mass spectrometry technology, and identified high-toxicity substances are determined by virtue of a quantitative structure-activity relation model; and finally, constructing a pretreatment and instrument detection method so as to realize accurate quantitative analysis of the key toxic substances. According to the method, the recognition efficiency of high-toxicity new disinfection byproducts in drinking water is remarkably improved, a toxicity evaluation system is perfected, the quantitative detection process is optimized, rapid batch analysis is achieved, efficient technical support is provided for drinking water safety guarantee, and the requirements of modern drinking water supervision for wide coverage and high-frequency monitoring are met.

Technologies for quantitative structure-activity relationship (QSAR) modeling for lipid nanoparticle (LNP) biotherapeutics include a computing device that receives a training data set including LNP test results, which each include an LNP chemical formulation and a corresponding result value of an LNP target variable, such as activity or cytotoxicity. The computing device extracts multiple input features for each LNP chemical formulation, where each of the input features is indicative of an attribute of a component or a composition of the LNP chemical formulation. The computing device trains a machine learning model to predict the LNP target variable with the input features and the result values of the training data set. A computing device may predict the LNP target variable result by extracting input features from a supplied LNP chemical formulation and supplying the input features to the trained machine learning model. Other embodiments are described and claimed.

This invention belongs to the field of detection technology and relates to a highly sensitive method for quantitative analysis and structural identification of cobalamin, as well as its application. Using three common natural cobalamins with different structures found in the environment as standards, a full-scan, automatically triggered secondary mass spectrometry acquisition method is employed to obtain the cobalamin precursor ion and a series of high-abundance characteristic ion fragments. Based on the principle that fragments of the same cobalamin have the same fragment ion m / z and the fragment ion search function of the software, the structure of unknown cobalamin is identified. Then, a parallel reaction monitoring mass spectrometry quantitative method is used. A large number of interfering ions are filtered out by a quadrupole, and the cobalamin low-position ligand fragment ions with the highest abundance and high identifiability are used as quantitative ions to analyze the type and content of cobalamin in the sample. This invention, by combining two data acquisition methods of high-resolution mass spectrometry, can effectively eliminate interference from the endogenous matrix, improve the sensitivity of sample analysis, and enhance the specificity of quantification, thereby achieving highly sensitive detection of different types of cobalamin.

The application discloses a kind of temperature-resistant water-soluble monomer design methods based on generative artificial intelligence.The method comprises the following steps: constructing polymer thermal performance dataset, training quantitative structure-property relationship prediction model;Adopt generative artificial intelligence strategy, by the structure deconstruction and recombination of existing temperature-resistant monomer and water-soluble monomer, generate massive virtual monomer library;Build water-soluble evaluation index considering solubility parameter, oil-water partition coefficient and hydrogen bond number;Combined with synthesis accessibility score, multi-objective high-throughput screening is carried out on virtual monomer library.The application realizes the synergistic optimization and high-throughput screening of the temperature resistance and water solubility of candidate monomers, solves the problems of limited types of existing temperature-resistant water-soluble monomers, mainly relying on experience trial and error in research and development, and low screening efficiency, especially the lack of unified and calculable evaluation index for water solubility, which makes it difficult to optimize and screen the temperature resistance and water solubility as constraints in the process of artificial intelligence aided design.

The invention discloses a method and a mathematical model for high-precision prediction of solubility parameters of a pure compound which is composed of 12 or less elements of hydrogen, carbon, nitrogen, oxygen, sulfur, fluorine, chlorine, bromine, iodine, silicon, phosphorus, arsenic and the like and has the atom number of 25 or less (excluding hydrogen). Wherein the model, as a universal quantitative structure-property relationship model, is obtained by determining an optimal model from a plurality of multiple linear regression models by means of a stepwise selection method, and the model takes some descriptors in various molecular descriptors as independent variables, takes solubility parameters as dependent variables, and takes the dissolvability parameters as dependent variables. The value of the molecule descriptor included in the model can be input in a short time and the solubility parameter can be output, so that the solubility parameter of a compound purely composed of the molecule can be predicted for any molecule as long as the specific value of the molecule descriptor included in the model is known.

The application relates to a method for constructing a quantitative structure-activity relationship of a dendrimer molecular structure and salt tolerance and application. According to the quantitative structure-activity relationship, the salt tolerance of an oil field chemical polymer can be obtained according to the molecular structure of the dendrimer, without needing to pass through traditional experiments or complicated molecular simulation dynamics calculation. Meanwhile, the quantitative structure-activity relationship can be used to quickly screen the dendrimer with good salt tolerance, and has a good reference significance for the rapid design and research and development of the salt-resistant polymer for drilling fluid and even oil field chemicals.

Provided is a mathematical model with which it is possible to accurately predict the standard boiling point of a pure compound having 25 or less atoms (excluding hydrogen), said pure compound comprising 12 or less elements such as hydrogen, carbon, nitrogen, oxygen, sulfur, fluorine, chlorine, bromine, iodine, silicon, phosphorus, and arsenic. Wherein the model, as a universal quantitative structure-property relationship model, is obtained by determining an optimal model from a plurality of multiple linear regression models by means of a stepwise selection method, and the model takes some of the various molecular descriptors as independent variables, takes a standard boiling point as a dependent variable, and takes the standard boiling point as a dependent variable; according to the present invention, the values of the molecule descriptors included in the model can be received and input in a short time and the standard boiling point can be output, so that the standard boiling point of the compound purely composed of the molecule can be predicted for any molecule satisfying the requirements of the present invention as long as the specific values of the molecule descriptors included in the model are known.

1. Name of the design product: Graphical user interface for creating tasks for quantitative structure-activity analysis of electronic devices. 2. Purpose of this design product: an electronic device. 3. The key design point of this design product lies in the interface content of the graphical user interface. 4. The picture or photo that best illustrates the key points of the design: main view. 5. Conventional design, omitting rear view, left view, right view, top view and bottom view. 6. Purpose of the graphical user interface: It is used to create training tasks and prediction tasks in the process of quantitative structure-activity analysis of the correlation between chemical structure and biological activity. The main view is the user task management interface, which manages the prediction tasks and training task sets created by the user. Click New Training Task in the upper right corner of the main view to enter the interface change state Figure 1, that is, the training task creation interface. The user clicks the Please upload file button to upload the file and enters the interface change state Figure 2. Create a form and select smiles_col. Enter the interface change state Figure 3. After the user selects the target_cols option, enter the interface change state Figure 4. Click the Validate Data button in the form to validate the data and enter the interface change state Figure 5. Click Submit Task at the bottom of the form to submit the task and enter the interface change state Figure 6. Click the New Prediction Task button in the upper right corner of the interface to enter the prediction task creation interface. Interface change state Figure 7. Click the Please upload file button to upload the file and enter the interface change state Figure 8. After filling in the smiles_col option in the form, enter the interface change state Figure 9. Click the Validate Data button to validate the data set and enter the interface change state Figure 10. Click Choose Training at the bottom Model button, a training model list pops up on the right side of the interface, and the interface change status diagram 11 is displayed. After selecting a training model, the interface change status diagram 12 is displayed. After clicking the Confirm button in the lower right corner to confirm, the interface change status diagram 13 is displayed. Click the Submit button at the bottom of the form, the page jumps to the user task management interface, and the interface change status diagram 14 is displayed.

The present invention provides a method for calculating the reaction rate constant of hydroxyl radical with dissolved organic matter in water, which can quickly and accurately calculate the reaction rate constant of hydroxyl radical with dissolved organic matter in water. A method for predicting the reaction rate constant of hydroxyl radical with dissolved organic matter in water by using a quantitative structure-activity relationship model is also disclosed. Each process of model establishment and verification in the present invention strictly complies with the OECD model construction and use guidelines. The obtained model has a high goodness of fit, good robustness and strong predictive ability. The instruments required for the determination of descriptors involved in the prediction model established in the present invention are mainly ultraviolet-visible spectrophotometer, TOC analyzer, three-dimensional fluorescence spectroscopy, etc. These instruments have been widely used to characterize DOM due to their advantages of high sensitivity, easy processing and analysis, and low cost. The selected descriptors have strong mechanistic interpretability. The obtained model has a simple form, good transparency and is easy to be programmed and popularized.

The present application relates to the field of environmental science and computational toxicology, and specifically provides a method for constructing species sensitivity distribution and performing soil ecological risk probability assessment based on multi-species quantitative structure-activity relationship model, which constructs QSAR model for each receptor and combines Monte Carlo simulation technology, models three key uncertainty sources of model application domain confidence, model reliability factor and ecological weight as Beta distribution, Logit- Normal distribution and truncated normal distribution respectively, randomly samples and calculates comprehensive weight in each iteration, constructs SSD curve based on weighted toxicity data, and then obtains probability distribution of hazard concentration, and finally, outputs risk probability distribution of pollutants according to comparison between pollutant concentration and HC5 distribution in the environment, which quantitatively the uncertainty of QSAR model and ecological parameters, realizes the transformation from point estimation to probability distribution, and provides a scientific basis with statistical significance for soil ecological risk classification management and remediation decision.

The invention discloses a mathematical model capable of predicting the flammability lower limit temperature of a pure compound which is composed of 12 or less elements such as hydrogen, carbon, nitrogen, oxygen, sulfur, fluorine, chlorine, bromine, iodine, silicon, phosphorus, arsenic and the like and has the atom number of 25 or less (excluding hydrogen) with high accuracy. Wherein the model is obtained as a universal quantitative structure-property relationship model by determining an optimal model from a plurality of multiple linear regression models by means of a stepwise selection method, and the model takes some descriptors in various molecular descriptors as independent variables, takes flammability lower limit temperature as a dependent variable, and takes the flammability lower limit temperature as a dependent variable. According to the present invention, the value of the molecule descriptor included in the input model can be received in a short time, and the flammability lower limit temperature can be output, such that the flammability lower limit temperature of the compound singly composed of the molecule can be predicted for any molecule satisfying the requirements of the present invention as long as the specific value of the molecule descriptor included in the model is known.

A quantitative structure-activity relationship model construction method based on model robustness relates to a quantitative structure-activity relationship model construction method, in particular to a quantitative structure-activity relationship local model construction method based on model robustness and a consistency model construction method of the quantitative structure-activity relationship local model, taking a compound liver metabolic clearance rate prediction model as an object, belonging to the field of chemical information science and bioinformatics. In view of the fact that the current global modeling method is difficult to cope with the complex modeling compounds of the liver metabolic clearance rate prediction model, the model robustness is poor, and the prediction performance is difficult to further improve, the QSAR local model based on the activity mechanism is difficult to be applied in practice, and the QSAR local model based on the chemical structure similarity lacks the technical means to guarantee the model robustness, the application provides a QSAR local model construction method based on model robustness, and a support vector machine regression technology is used to establish a QSAR local model.

Provided is a mathematical model with which it is possible to accurately predict the critical volume of a pure compound having 25 or less atoms (excluding hydrogen), said pure compound comprising 12 or less elements such as hydrogen, carbon, nitrogen, oxygen, sulfur, fluorine, chlorine, bromine, iodine, silicon, phosphorus, and arsenic. Wherein the model is obtained as a universal quantitative structure-property relationship model by determining an optimal model from a plurality of multiple linear regression models by means of a step-by-step selection method, and the model takes some of the various molecular descriptors as independent variables, takes critical volumes as dependent variables, and takes the critical volumes as dependent variables. According to the present invention, the value of the molecule descriptor included in the input model can be received in a short time and the critical volume can be output, such that the critical volume of the compound purely composed of the molecule can be predicted for any molecule satisfying the requirements of the present invention as long as the specific value of the molecule descriptor included in the model is known.

The invention belongs to the technical field of artificial intelligence, and provides a drug research and development architecture and method based on artificial intelligence, electronic equipment and a computer program.The drug research and development architecture constructs a multi-artificial-intelligence-model cluster interactive agent, and applies a large language model to deeply learn and integrate drug related data for new drug research and development. Fusing multi-dimensional information of the whole process of new drug research and development, performing effective analysis, and performing global reasoning to generate potential novel drug candidate molecules and biomolecular targets corresponding to the potential novel drug candidate molecules; and interaction, iteration and optimization are carried out through the thinking chain and various types of artificial intelligence models. The artificial intelligence models comprise a molecular docking model, a protein structure prediction model, a multi-omics conjoint analysis model, a pharmacokinetic model and a quantitative structure-activity model, so that the current situations of high failure rate and high risk of drug candidate molecules and targets in later chemical experiments, biological experiments and clinical experiments are effectively avoided; and the efficiency and success rate of drug research and development are greatly improved.